A state monitoring system

专利摘要:
A method of operating a device for analyzing the state of a machine having a part rotating at a rotational speed (fROT), comprising the steps of: receiving a first digital signal (SRED, SMD, SENV) due to mechanical vibrations originating in rotation of said part; analyzing the first digital signal (SRED, SMD, SENV) to detect amplitude peaks (Ap) during a finite time period (Pm), the fi rivet time period corresponding to a certain amount (R) of rotation of the rotatable member; wherein the determined amount (R) of rotation corresponds to more than one revolution of the monitored rotatable member; defining a plurality of (NR) amplitude ranges; sort the detected amplitude peak values (Ap) into corresponding amplitude ranges so that the occurrence (N) of detected amplitude peak values (Ap) within said plurality of amplitude ranges is represented; estimate a representative amplitude peak value (APR) depending on the sorted amplitude peak values (Ap) and the determined amount (R).
公开号:SE1000631A1
申请号:SE1000631
申请日:2010-06-11
公开日:2011-07-19
发明作者:Lars-Olov Elis Hedin
申请人:Spm Instr Ab；
IPC主号:

专利说明:

The device described in 2 has a sensor for generating a measured value indicating vibration at a measuring point. The device described in WO 03062766 has a data processor and a memory. The memory can store program code which, when run on the data processor, will cause the analyzer to perform a machine condition monitoring function. Such a machine condition monitoring function may include shock pulse measurement.
Summary One aspect of the invention relates to the problem of providing an improved method and apparatus for analyzing the condition of a machine having a rotating member.
This problem is addressed by a method of operating a device for analyzing the state of a machine having a part rotating at a rotational speed (fRoT), comprising the steps of: receiving a first digital signal (SRED, SMD, SENV) which due to mechanical vibrations originating in rotation of the part; analyzing the first digital signal (SRED, SMD, SENV) to detect amplitude peaks (Ap) during a finite time period (Pm), the finite time period corresponding to a certain amount (R) of rotation of the rotatable part; said certain amount (R) of rotation corresponding to more than one revolution of the rotatable part; defining a plurality of (NR) amplitude ranges; sort the detected amplitude peaks (Aβ) into corresponding amplitude ranges so that the presence (N) of detected amplitude peaks (Aβ) within said amplitude range is indicated; estimate a representative amplitude peak value (APR) depending on the sorted amplitude peak values (Ap) and said certain amount (R).
BRIEF DESCRIPTION OF THE DRAWINGS For a simple understanding of the present invention, it will be described by way of example and with reference to the accompanying drawings in which: Fig. 1 shows a schematic block diagram of an embodiment of a conditioning system 2 according to an embodiment of the invention.
Fig. 2A is a schematic block diagram of an embodiment of a portion of the state analysis system 2 shown in Fig. 1.
Fig. 2B is a schematic block diagram of an embodiment of a sensor interface.
Fig. 2C is an illustration of a measurement signal from a vibration sensor.
Fig. 2D illustrates a measurement signal amplitude generated by a shock pulse sensor.
Fig. 2E illustrates a measurement signal amplitude generated by a vibration sensor.
Fig. 3 is a simplified illustration of a shock pulse measurement sensor according to an embodiment of the invention.
Fig. 4 is a simplified illustration of an embodiment of the memory 60 and its contents.
Fig. 5 is a schematic block diagram of an embodiment of the analysis apparatus at a customer site with a machine 6 having a movable shaft.
Fig. 6A illustrates a schematic block diagram of an embodiment of the preprocessor according to an embodiment of the present invention.
F ig. 6B illustrates a schematic block diagram of an embodiment of the preprocessor comprising a digital rectifier.
Fig. 7 illustrates an embodiment of the evaluator.
Fig. 8 is a schematic illustration of a rectified signal that could be supplied by the rectifier shown in Fig. 6B.
Fig. 9 illustrates a histogram that is the result of a measurement.
Fig. 10 illustrates a histogram resulting from another measurement.
Fig. 11A is a flow chart illustrating an embodiment of a method of operating the device so that it is prepared to perform a peak value state analysis.
Fig. 11B is a flow chart illustrating an embodiment of a method of operating the device so that it performs peak value state analysis.
Fig. 12A is a flow chart illustrating an embodiment of a method for performing a peak measurement session.
Fig. 12B is a flow chart illustrating an embodiment of a method for performing a peak value measurement session and addressing the influence of bursts of noise amplitude peaks.
Fig. 13A illustrates a histogram with a plurality of amplitude compartments.
Fig. 13B is a schematic illustration of a plurality of memory positions arranged as a table. Fig. 13C is an illustration of a cumulative histogram table corresponding to the histogram table of Fig. 13B.
Fig. 14A is a flow chart illustrating an embodiment of a method for determining a representative amplitude peak value based on the amplitude peak values Aβ collected during the measurement session.
Fig. 14B is a flow chart illustrating a further embodiment of a method for estimating a representative amplitude peak value APR based on the amplitude peak values Aβ collected during the measurement session.
Fig. 15A is an illustration reflecting the principle of a cumulative histogram resulting from a measurement.
Detailed Description of Embodiments In the following description, similar features in different embodiments may be indicated by the same reference numerals.
Fig. 1 shows a schematic block diagram of an embodiment of a state analysis system 2 according to an embodiment of the invention. Reference numeral 4 denotes a customer site with a machine 6 having a moving part 8. The moving part may comprise bearings 7 and a shaft 8 which, when the machine is in operation, rotates. The operating state of the shaft 8 or of the bearing 7 can be determined depending on vibrations originating from the shaft and / or the bearing when the shaft rotates. The customer site 4, which can also be referred to as the customer part or the user part, can for example be the site of a wind farm, i.e. a group of wind turbines at a site, or the site of a paper mill, or any other manufacturing plant having machines with moving parts.
An embodiment of the condition analysis system 2 is operational when a sensor 10 is attached to or at a measuring point 12 on the housing of a machine 6. Although Fig. 1 illustrates only two measuring points 12, it should be understood that a place 4 can comprise any number of measurements. points 12. The condition analysis system 2 shown in Fig. 1 comprises an analysis device 14 for analyzing the condition of a machine on the basis of measured values supplied by the sensor 10. The analysis device 14 has a communication port 16 for bidirectional data communication. The communication port 16 is connectable to a communication network 18, for example via a data interface 19. The communication network 18 may be the worldwide Internet, also known as the Internet. The communication network 18 may also include a public circuit switched telephone network.
A server computer 20 is connected to the communication network 18. The server 20 may include a database 22, I / O user interface 24 and data processing hardware 26, and a communication port 29. The server computer 20 is located at a location 28, which is geographically separate from the customer location 4. The server location 28 may be in a first city, such as the Swedish capital Stockholm, and the customer site may be in another city, such as Stuttgart, Germany or Detroit in Michigan, USA. Alternatively, the server location 28 may be in a first part of a city and the customer location may be in another location of the same city. The server location 28 can also be referred to as supplier part 28, or supplier part location 28.
According to an embodiment of the invention, a central control site 31 comprises a control computer 33 which has data processing hardware and software for monitoring a plurality of machines at the customer site 4. The machines 6 may be wind turbines or gearboxes used in wind turbines. Alternatively, the machines may include machinery in, for example, a paper mill. The control computer 33 may comprise a database 22B, input / output user interface 24B and data processing hardware 26B, and a communication port 29B. The central control location 31 can be separated from the customer location 4 by a geographical distance. By means of the communication port 29B, the control computer 33 can be connected to communicate with the analysis device 14 via port 16. The analysis device 14 can deliver measurement data which is partly processed so that further signal processing and / or analysis can be performed at the central location 31 of the control computer 33.
A supplier company occupies the supplier part site 28. The supplier company can sell and deliver analysis devices 14 and / or software for use in an analysis device 14. The supplier company can also sell and deliver analysis software for use in the control computer at the central control site 31. Such analysis software 94, 105 is discussed in in connection with Fig. 4 below. Such analysis software 94, 105 can be delivered via transmission over the communication network 18. According to an embodiment of the system 2, the device 14 is a portable device which can be connected to the communication network 18 from time to time.
According to another embodiment of the system 2, the device 14 is connected to the communication network 18 substantially continuously. Thus, the device 14 according to this embodiment can essentially always be "on-line-available" for communication with the supplier computer 20 and / or with the control computer 33 at control location 31.
Fig. 2A is a schematic block diagram of an embodiment of a portion of the state analysis system 2 shown in Fig. 1. The state analysis system, as illustrated in Fig. 2A, comprises a sensor unit 10 for generating a measured value. The measured value can be dependent on movement or, more precisely, dependent on vibrations or shock pulses caused by bearings when the shaft rotates.
An embodiment of the condition analysis system 2 is operable when a device 30 is fixedly mounted on or at a measuring point of a machine 6. The device 30 mounted at the measuring point may be called a stud 30. A stud 30 may comprise a connection coupling 32 to which the sensor unit 30 is releasably attachable. The connection coupling 32 may, for example, comprise double starting threads in order to enable the sensor unit to be mechanically coupled to the stud by means of rotation of one revolution.
A measuring point 12 may comprise a threaded recess in the housing of the machine. A stud 30 may have a protruding piece with threads corresponding to the threads of the recess to enable the stud to be firmly attached to the measuring point by inserting the stud into the recess as a bolt.
Alternatively, a measuring point 12 may comprise a threaded recess in the housing of the machine, and the sensor unit 10 may comprise corresponding threads so that it can be inserted directly into the recess. Alternatively, the measuring point is marked on the machine housing only by means of a painted mark. 10 15 20 25 30 7 The machine 6, which is exemplified in Fig. 2, may have a rotating shaft with a certain shaft diameter d1. The shaft of the machine 24 can rotate at a certain rotational speed V1 when the machine 6 is in use.
The sensor unit 10 may be connected to the device 14 for analyzing the condition of the machine. Referring to Fig. 2A, the analyzer 14 includes a sensor interface 40 for receiving a measurement signal or measurement data generated by the sensor 10. The sensor interface 40 is coupled to a data processing means 50 which can control the operation of the analyzer 14 in accordance with program code. The data processing means 50 is also connected to a memory 60 for storing the program code.
According to an embodiment of the invention, the sensor interface 40 comprises an input 42 for receiving an analog signal, the input 42 being connected to an analog-to-digital converter 44 (A / D converter 44), the digital output of which is connected to a data controller. the operating means 50. The A / D converter 44 samples the received analog signal with a certain sampling frequency fs to supply a digital measurement data signal SMD having a certain sampling frequency fs and the amplitude of each sample depending on the amplitude of the received analog signal at the sampling time. .
According to another embodiment of the invention, illustrated in Fig. 2B, the sensor interface 40 comprises an input 42 for receiving an analog signal SEA from a shock pulse measurement sensor, a conditioning circuit 43 coupled to receive the analog signal, and an A / D converter 44 coupled to receive the conditioned analog signal from the conditioning circuit 43. The A / D converter 44 samples the received conditioned analog signal with a certain sampling frequency fs so as to supply a digital measurement data signal SMD having a certain sampling frequency fs and wherein the amplitude of each sample depends on the amplitude of the received analog signal at the time of sampling.
The sampling theorem guarantees that band-limited signals (i.e. signals having a maximum frequency) can be reconstructed perfectly from their sampled versions, if the sampling frequency fs is more than twice the maximum frequency fSEAmaX of the analog signal SEA to be monitored. The frequency which is equal to half of the sampling frequency is therefore a theoretical limit for the highest frequency which can be unambiguously represented by the sampled signal SMD. This frequency (half the sampling frequency) is called the Nyquist frequency in the sampled system. Frequencies above the Nyquist frequency fN can be observed in the sampled signal, but their frequency is unambiguous.
This means that a frequency component with frequency f can not be distinguished from other components with frequencies B * fN + f, and B * fN - f for integers B other than zero. This ambiguity, known as aliasing, can be handled by filtering the signal with an anti-aliasing filter (usually a low-pass filter with a cut-off frequency close to the Nyquist frequency) before conversion to the sampled discrete representation.
To provide a safety margin in terms of allowing a non-ideal filter to have a certain edge in the frequency response, the sampling frequency can be selected to a value higher than 2.
Thus, according to embodiments of the invention, the sampling frequency can be set to fs = k * fsEAmax where k is a factor having a value higher than 2.0 Thus, the factor k can be selected to a value higher than 2.0. Preferably, the factor k can be selected to a value between 2.0 and 2.9 in order to provide a good safety margin and at the same time avoid generating unnecessarily many sample values. According to one embodiment, the factor k is advantageously selected so that 100 * k / 2 gives an integer. According to one embodiment, the factor k can be set to 2.56. Selecting k to 2.56 gives the result that 100 * k = 256 = 2 raised to 8.
According to one embodiment, the sampling frequency fs of the digital measurement data signal SMD can be fixed to a certain value fs, such as for example fs = 102 kHz.
Thus, when the sampling frequency fs is fixed to a certain value fs, the maximum frequency fSEAmaX of the analog signal SEA will be: fSEAmax = fS l k 10 15 20 25 30 9 where fSEAmaX is the highest frequency to be analyzed in the sampled signal.
Thus, when the sampling frequency fs is fixed to a certain value fs = 102 400 Hz, and the factor k is set to 2.56, the maximum frequency fSEAmax of the analog signal SEA will be: fSEAmaX = fs / k = 102 400/2 , 56 = 40 kHz Thus, a digital measurement data signal SMD is generated, with a certain sampling frequency fs, depending on the received analog measurement signal SEA. The digital output 48 of the AID converter 44 is connected to the data processing means 50 via an output 49 of the sensor interface 40 so as to supply the digital measurement data signal SMD to the data processing means 50.
The sensor unit 10 may comprise a vibration signal converter, the sensor unit being designed to be physically connected to the connection connection of the measuring point so that machine vibrations at the measuring point are transmitted to the vibration signal converter. According to an embodiment of the invention, the sensor unit comprises a signal converter which has a piezoelectric element. When the measuring point 12 vibrates, the sensor unit 10, or at least a part of it, will also vibrate and the signal converter then generates an electrical signal whose frequency and amplitude depend on the mechanical vibration frequency resp. the vibration amplitude at the measuring point 12. According to an embodiment of the invention, the sensor unit 10 is a vibration sensor which provides an analog amplitude signal of, for example, 10 mV / g in the frequency range 1.00-10 000 Hz. Such a vibration sensor is designed to deliver substantially the same amplitude of 10 mV regardless of whether it is subjected to acceleration 1 g (9.82 m / sz) at 1 Hz, 3 Hz or 10 Hz.
Thus, a typical vibration sensor has a linear response within a specified frequency range up to about 10 kHz. Mechanical vibrations in the frequency range, which originate in rotating machine parts, are usually caused by imbalance or misalignment.
However, when mounted on a machine, an linear response vibration sensor typically also has a plurality of different mechanical resonant frequencies that depend on the physical path between the sensor and the vibration source. 10 15 20 25 30 10 An injury in a roller bearing causes relatively sharp electric waves, known as shock pulses, which move along a physical path in the housing of the machine before they reach the sensor. Such shock pulses often have a wide frequency spectrum. The amplitude of a roller bearing shock pulse is typically lower than the amplitude of a vibration caused by imbalance or misalignment.
The wide frequency spectrum of the shock pulse signatures enables them to activate a "ringing response" or a resonance at a resonant frequency associated with the sensor.
Thus, a typical measurement signal from a vibration sensor may have a waveform shown in Fig. 2C, i.e. a dominant low frequency signal with a superimposed more high frequency resonant "ringing response" with lower amplitude.
To enable analysis of the shock pulse signature, which often originates in a bearing damage, the low-frequency component must be filtered out. This can be achieved by means of a high-pass filter or by means of a band-pass filter. However, these filters must be adjusted so that the low frequency signal part is blocked while the high frequency signal part is delivered further. An individual vibration sensor will typically have a resonant frequency associated with the physical path from one shock pulse signal source, and another resonant frequency associated with the physical path from another shock pulse source, as mentioned in US 6,053,047. Filter adjustment for further delivery of the high frequency signal portion requires thus individual adaptation when a vibration sensor is used.
When such a filter is properly adjusted, the resulting signal will consist of the shock pulse signature (s). However, the analysis of the shock pulse signature (s) resulting from a vibration sensor is somewhat impaired by the fact that the amplitude response as well as the resonant frequency by their nature varies depending on the individual physical path from the shock pulse signal sources.
Advantageously, these disadvantages associated with vibration sensors can be remedied by using a shock pulse measurement sensor. The shock pulse measurement sensor is designed and adapted to provide a predetermined mechanical resonant frequency as described in more detail below.
This property of the shock pulse measurement sensor advantageously provides repeatable measurement results in that the output signal from a shock pulse measurement sensor has a stable resonant frequency substantially independent of the physical path between the shock pulse signal source and the shock pulse sensor. Furthermore, mutually different individual shock pulse sensors provide very small, if any, deviations in resonant frequency.
An advantageous effect of this is that the signal processing is simplified, in that filters do not have to be adjusted individually, in contrast to the cases described above when vibration sensors are used. Furthermore, the amplitude response from shock pulse sensors is well defined so that an individual measurement provides reliable information when the measurement is performed in accordance with appropriate measurement methods defined by S.P.M. Instrument AB.
Fig. 2D illustrates a measurement signal amplitude generated by a shock pulse sensor, and Fig. 2E illustrates a measurement signal amplitude generated by a vibration sensor. Both sensors have been subjected to the same series of mechanical shocks without the typical low frequency signal content.
As can be clearly seen in Figs. 2D and 2E, the duration of a resonant response to a shock pulse signature from the shock pulse sensor is shorter than the corresponding resonant response to a shock pulse signature from the vibration sensor.
The property of this shock pulse measurement sensor to provide distinct shock pulse signature responses has the beneficial effect of providing a measurement signal from which it is possible to distinguish between different mechanical shock pulses which occur within a short period of time.
According to an embodiment of the invention, the sensor is a shock pulse measurement sensor. Fig. 3 is a simplified illustration of a shock pulse measurement sensor 10 according to an embodiment of the invention. According to this embodiment, the sensor comprises a part 110 having a certain mass or weight and a piezoelectric element 120. The piezoelectric element 120 is somewhat flexible so that it can be compressed and expanded when subjected to external force. The piezoelectric element 120 is provided with electrically conductive bearings 130 resp. 140 on opposite surfaces. As the piezoelectric element 120 shrinks and expands, it generates an electrical signal which is captured by the conductive layers 130 and 140. Thus, a mechanical vibration is converted into an analog electrical measurement signal SEA, which is supplied on the output signal terminals 145, 150. The piezoelectric element 120 can be placed between the weight 110 and a surface 160 which, during operation, is physically attached to the measuring point 12, as illustrated in Fig. 3. The shock pulse measuring sensor 10 has a resonant frequency which depends on the mechanical characteristics of the sensor, such as the mass m of the weight part 110 and the elasticity of the piezoelectric element 120. Thus, the piezoelectric element has an elasticity and a spring constant k. The mechanical resonant frequency fRM of the sensor is therefore dependent on the mass m and the spring constant k.
According to one embodiment of the invention, the mechanical resonant frequency fRM of the sensor can be determined by the following equation: fRM = 1 / (2n) / (k / m) (eq1) According to another embodiment, the actual mechanical resonant frequency of a shock pulse measurement sensor 10 may also depend on other factors. , such as how the sensor is attached to the housing 6 of the machine.
Thus, the resonant shock pulse measurement sensor 10 is particularly sensitive to vibrations having a frequency at or near the mechanical resonant frequency fRM. The shock pulse measuring sensor 10 can be designed so that the mechanical resonant frequency fRM is somewhere in the range from 28 kHz to 37 kHz. According to another embodiment, the mechanical resonant frequency fRM is somewhere in the range from 30 kHz to 35 kHz.
Thus, the analog electrical measurement signal has an electrical amplitude that can vary across the frequency spectrum. In order to describe the theoretical background, it can be assumed that if the shock pulse measurement sensor 10 were subjected to mechanical vibrations of identical amplitude within all frequencies from, for example, 1 Hz to, for example, 200,000 kHz, the amplitude of the analog signal SEA from the shock pulse measurement sensor will have a maximum. at the mechanical resonant frequency fRM, since the sensor will resonate when "pushed" at that frequency.
Referring to Fig. 2B, the conditioning circuit 43 receives the analog signal SEA. The conditioning circuit 43 can be designed to be an impedance matching circuit arranged to adjust the input impedance of the A / D converter as perceived by the sensor connections 145, 150 so that an optimal signal transmission occurs. Thus, the conditioning circuit 43 can operate to match the input impedance Z ", viewed from the sensor terminals 145, 150, so that a maximum electrical power is supplied to the AID converter 44. According to one embodiment of the conditioning circuit 43, the analog the signal SEA to a primary winding of a transformer, and a conditioned analog signal is supplied by a secondary winding of a transformer.The primary winding has n1 turns and the secondary winding has n2 turns, the ratio n1 / n2 = m2. from the conditioning circuit 43. The A / D converter 44 has an in-impedance 244, and the in-impedance of the A / D converter viewed from the sensor terminals 145, 150 will be (n1 / n2) 2 * 244, when the conditioning circuit 43 is connected between the sensor connections 145, 150 and the input connections of the A / D converter 44.
The A / D converter 44 samples the received conditioned analog signal with a certain sampling frequency fs so as to supply a digital measurement data signal SMD with a certain sampling frequency fs where the amplitude of each sample depends on the amplitude of the received analog signal at the sampling time.
According to embodiments of the invention, the digital measurement data signal SMD is supplied to a means 180 for digital signal processing (see Fig. 5).
According to one embodiment of the invention, the digital signal processing means 180 comprises the data processor 50 and program code for causing the data processor 50 to perform digital signal processing. According to an embodiment of the invention, the processor 50 is constituted by a digital signal processor. The digital signal processor can also be called DSP.
Referring to Fig. 2A, the data processing means 50 is connected to a memory 60 for storing the program code. The program memory 60 is preferably a non-volatile memory.
The memory 60 can be a read / write memory, i.e. enabling both reading of data from the memory and writing of new data to the memory 60. According to one embodiment, the program memory 60 consists of a FLASH memory. The program memory 60 may include a first memory segment 70 for storing a first set of program code 80 that is executable so that it controls the analyzer 14 to perform basic functions (Fig. 2A and Fig. 4).
The program memory may also include a second memory segment 90 for storing a second set of program code 94. The second set of program code 94 in the second memory segment 90 may include program code for causing the analyzer to process the detected signal, or signals. , so that it generates a preprocessed signal or a set of preprocessed signals. The memory 60 may also include a third memory segment 100 for storing a third set of program code 104. A set of program codes 104 in the third memory segment 100 may include program code for causing the analyzer to perform a selected analysis function 105. When the analysis function is performed, it may cause the analyzer to present a corresponding analysis result on the user interface 106 or to deliver the analysis result on port 16 (see Fig. 1 and Fig. 2A and Fig. 7).
The data processing means 50 is also connected to the read / write memory 52 for data storage. Further, the data processing means 50 may be coupled to an analyzer communication interface 54. The analyzer communication interface 54 provides bidirectional communication with a measuring point communication interface 56 which is attached only to or near the measuring point of the machine.
The measuring point 12 may comprise a connection connection 32, a readable and writable information carrier 58, and a measuring point communication interface 56.
The writable information carrier 58, and the measurement point communication interface 56 may be provided in a separate device 59 located in the vicinity of the study 30, as illustrated in Fig. 2. Alternatively, the writable information carrier 58, and the measuring point communication interface 56 may be provided within the study 30. This is described in greater detail. detail in WO 98/01831, the contents of which are hereby incorporated by reference.
The system 2 is arranged to allow bidirectional communication between the measuring point communication interface 56 and the analyzer communication interface 54.
The measuring point communication interface 56 and the analyzer communication interface 54 are preferably designed to allow wireless communication. According to one embodiment, the measuring point communication interface and the analyzer communication interface are designed to communicate with each other via radio frequency (RF) signals. This embodiment includes an antenna in the measurement point communication interface 56 and another antenna in the analyzer communication interface 54. Fig. 4 is a simplified illustration of an embodiment of the memory 60 and its contents.
The purpose of the simplified illustration is to convey an understanding of the general idea of storing various program functions in memory 60, and it is not necessarily an accurate technical description of the manner in which a program would be stored in a real memory circuit. The first memory segment 70 stores program code for controlling the analysis device 14 to perform basic tasks. Although the simplified illustration of Fig. 4 shows pseudocode, it is to be understood that the program code 80 may be machine code, or program code at any level that may be executed or interpreted by the data processor 50 (Fig. 2A).
The second memory segment 90, illustrated in Fig. 4, stores a second set of program code 94. When executed on the data processing means 50, the program code 94 in segment 90 will cause the analysis device 14 to perform a function, such as a digital signal processing function. The function can include an advanced mathematical processing of the digital measurement data signal SMD. According to embodiments of the invention, the program code 94 is arranged to cause the processor means 50 to perform signal processing functions described in connection with Figs. 5, 6, 9 and / or Fig. 16 of this document.
As described above in connection with Fig. 1, a computer program for controlling the operation of the analyzer may be downloaded from the server computer 20. This means that the program-to-be-downloaded is transmitted over the communication network 18. This can be done by modulating a carrier to carry the program over the communication network 18. Thus, the downloaded program can be loaded into a digital memory, such as the memory 60 (see Figs. 2A and 4). Thus, a signal processing program 94 and / or an analysis function program 104, 105 may be received via a communication port, such as port 16 (Figs. 1 & 2A), so that it is loaded into the memory 60. Similarly, a signal processing program 94 and / or an analysis function program 104, 105 is received via communication port 29B (Fig. 1) so that it is loaded into a program memory space in the computer 26B or in the database 22B.
One aspect of the invention relates to a computer program product, such as a program code means 94 and / or program code means 104, 105 which is loadable in a digital memory belonging to a device. The computer software product includes software code snippets for performing signal processing methods and / or analysis functions when said product is run on a data processing unit 50 in an apparatus for analyzing the state of a machine.
The expression "running on a data processing unit" means that the computer program plus the data processing unit performs a procedure of the kind described in this document.
The wording "a computer program product, loadable in a digital memory in a state analysis device" means that a computer program can be inserted into the digital memory in a state analysis device so as to obtain a state analysis device programmed to be capable of, or arranged to, perform a The term "loaded in a digital memory in a state analysis device" means that the state analysis device programmed in this way is capable of, or arranged to, perform a method of the kind described above.
The above-mentioned computer program product may also be loadable to a computer-readable medium, such as a compact disc or DVD. Such a computer-readable medium can be used for delivery of the program to a customer.
According to an embodiment of the analysis device 14 (Fig. 2A), it comprises a user input interface 102, with which an operator can interact with the analysis device 14. According to one embodiment, the user input interface 102 comprises a set of buttons 104. An embodiment of the analysis device 14 comprises a user output interface 106. The user output interface may include a display unit 106. When the data processing means 50 executes a basic program function provided in the basic program code 80, the data processing means 50 enables user interaction by means of the user input interface 102 and the display unit 106. The set of buttons 104 may be limited to a few buttons. for example, five buttons as illustrated in Fig. 2A. A central button 107 can be used for an ENTER or SELECT function, while other more peripheral buttons can be used to surface a mark on the display 106. In this way it should be understood that symbols and text can be entered into the device 14 via the user interface . The display unit 106 may, for example, display a number of symbols, such as the letters of an alphabet, while the cursor is movable on the display depending on user input so that the user is allowed to enter information. Fig. 5 is a schematic block diagram of an embodiment of the analysis device 14 at a customer site 4 with a machine 6 having a movable shaft 8. The sensor 10, which may be a shock pulse measurement sensor, is shown attached to a machine body 6 so that it captures mechanical vibrations and so that it delivers an analog measurement signal SEA indicative of the detected mechanical vibrations to the sensor interface 40. The sensor interface 40 may be constructed as described in connection with Fig. 2A or 2B. The sensor interface 40 delivers a digital measurement data signal SMD to a means 180 for digital signal processing.
The digital measurement data signal SMD has a sampling frequency fs, and the amplitude value of each sample depends on the amplitude of the received analog measurement signal SEA at the sampling time. According to one embodiment, the sampling frequency fs of the digital measurement data signal SMD can be fixed to a certain value fs, such as, for example, fs = 102 kHz.
The sampling frequency fs can be controlled by a clock signal supplied by a clock 190, as illustrated in Fig. 5. The clock signal can also be supplied to the digital signal processing means 180. The digital signal processing means 180 may generate information about the time duration of the received digital measurement data signal SMD depending on the received digital measurement data signal SMD, the clock signal and the relationship between the sampling frequency fs and the clock signal, since the duration between two consecutive sample values is equal to TS = 1. / fs.
According to embodiments of the invention, the digital signal processing means 180 comprises a preprocessor 200 for performing a preprocessing of the digital measurement data signal SMD to supply a preprocessed digital signal SMDp at an output 210. The output 210 is connected to an input 220 of an evaluator 230. The evaluator 230 is arranged to evaluate the pre-processed digital signal SMDp so that it delivers a result of the evaluation to a user interface 106. Alternatively, the result of the evaluation can be delivered to a communication port 16 so that transmission of the result is possible, for example, to a control computer 33 at a control position 31 (see Fig. 1).
According to one embodiment of the invention, the functions described in connection with the functional blocks in the digital signal processing means 180, preprocessor 200 and evaluator 230 may be provided by computer program code 94 and / or 104 described in connection with the memory areas 90 and 100 in connection with Figs. 4 above. 10 15 20 25 30 18 A user may need only a few basic monitoring functions to detect whether the condition of a machine is normal or abnormal. When detecting an abnormal condition, the user can call on specialized professional maintenance personnel to determine the exact nature of the problem, and to perform the necessary maintenance work. The professional maintenance staff often needs and uses a wide range of evaluation functions that make it possible to determine the nature of, and / or the cause of, an abnormal machine condition. Thus, different users of an analysis device 14 may place very different demands on the operation of the device. The term condition monitoring function is used in this document for a function for detecting whether the condition of a machine is normal or slightly deteriorated or abnormal.
The term condition monitoring function also includes an evaluation function that makes it possible to determine the nature of, and / or the cause of, an abnormal machine condition.
Examples of machine condition monitoring functions The condition monitoring functions F1, F2. Fn includes functions such as: vibration analysis, temperature analysis, shock pulse measurement, spectrum analysis of shock pulse measurement data, rapid Fourier transformation of vibration measurement data, graphical presentation of state data on a user interface, storage of state data in a writable information carrier on said state data in a writable information carrier in said device, tachometering, imbalance detection and error detection alignment.
According to one embodiment, the device 14 comprises the following functions: F1 = vibration analysis, F2 = shock pulse measurement, F3 = peak level measurement, F4 = spectrum analysis of shock pulse measurement data, F5 = fast Fourier transformation of vibration measurement data, F6 = graphical presentation of state data on a user interface of F7 state data in a writable information carrier on said machine, F8 = storage of state data in a writable information carrier 52 in said device, F9 = tachometer measurement, F10 = imbalance detection, and F11 = error correction detection, F12 = retrieval of state data from a writable information carrier 58 on the machine, F13 = performing top level measurement F3 and performing the function F12 "retrieving state data from a writable information carrier 58 on the machine" to enable a comparison or trend based on current top level measurement data and historical top level measurement data.
F14 = retrieving identification data from a writable information carrier 58 on the machine.
Embodiments of the function F7 "storage of state data in a writable information carrier on the machine", and F13 vibration analysis and retrieval of state data are described in more detail in WO 98/01831, the contents of which are hereby incorporated by reference.
The peak level analysis F3 can be performed on the basis of the enveloped time domain signal SENV supplied by envelope 250. The signal SENV is also called SMDP.
The peak level analysis F3 is adapted to monitor the signal for the duration of a peak level monitoring period PM in order to determine the maximum amplitude level.
The maximum amplitude level may be indicative of oil film thickness in a monitored layer.
Thus, the detected amplitude may be indicative of separation between metal surfaces in the scroll interface. The oil film thickness may depend on the supply of lubricant and / or on the alignment of the shaft. Furthermore, the oil film thickness can depend on the load on the shaft, ie on the force with which the metal surfaces are pressed against each other, the metal surfaces being, for example, a metal surface of a bearing and a metal surface of one shaft.
The actual detected value of the maximum amplitude level can also depend on the mechanical condition of the bearing surfaces, ie the condition of the bearing unit. Thus, the detected value of the maximum amplitude level may be due to roughness of the metal surfaces in the roll interface, and / or damage to a metal surface in the roll interface. The detected value of the maximum amplitude level may also depend on the presence of a loose particle in the storage unit.
Fig. 6A illustrates a schematic block diagram of an embodiment of the preprocessor 200 according to an embodiment of the present invention. In this embodiment, the digital measurement data signal SMD is coupled to a digital bandpass filter 240 having a lower cut-off frequency fLC, an upper cut-off frequency fUC and a passband bandwidth between the upper and lower cut-off frequencies.
The output of the digital bandpass filter 240 is connected to a digital envelope 250. According to one embodiment of the invention, the output of the envelope 250 is supplied to an output 260. The output 260 of the preprocessor 200 is connected to the output 210 of the digital signal processor 180 for delivery to the input 220 of the evaluator 230. .
The upper and lower cut-off frequencies of the digital bandpass filter 240 can be selected so that the SMD frequency components of the signal at the sensor resonant frequency fRM are within the passband bandwidth. As mentioned above, an amplification of the mechanical vibration is achieved by the sensor being mechanically resonant at the resonant frequency fRM. Thus, the analog measurement signal SEA represents an amplified value for the vibrations at and around the resonant frequency fRM. This means that the bandpass filter according to the embodiment according to Fig. 6 advantageously suppresses the signal at frequencies below and above the resonant frequency fRM, so that the components of the measuring signal at the resonant frequency fRM are further amplified. Advantageously, the digital bandpass filter 240 further limits the noise inherent in the measurement signal, since each noise component below the lower cut-off frequency fLC, and above the upper cut-off frequency fUC is also eliminated or attenuated. Thus, when using a resonant shock pulse measurement sensor 10 having a mechanical resonant frequency fRM in a range from a lowest resonant frequency value fRML to a highest resonant frequency value fRMU, the digital bandpass filter 240 can thus be constructed to have a lower cutoff frequency = fRML, and an upper cutoff frequency fRML. According to one embodiment, the lower cutoff frequency fLC = fRML = 28 kHz, and the upper cutoff frequency fUC = fRMU = 37 kHz. According to another embodiment, the mechanical resonant frequency fRM is somewhere in the range from 30 kHz to 35 kHz, and the digital bandpass filter 240 can then be designed to have a lower cut-off frequency fLC = 30 kHz and an upper cut-off frequency fUC = 35 kHz .
According to another embodiment, the digital bandpass filter 240 may be designed to have a lower cutoff frequency fLC that is lower than the lowest resonant frequency value fRM, and an upper cutoff frequency fUC that is higher than the highest resonant frequency value fRMU. For example, the mechanical resonant frequency fRM may be a frequency in the range from 30 kHz to 35 kHz, and the digital bandpass filter 240 may then be designed to have a lower cutoff frequency fLC = 17 kHz, and an upper cutoff frequency fUC = 36 kHz.
Thus, the bandpass filter 240 can deliver a digital passband measurement data signal SF with an advantageously low noise content and reflective mechanical vibrations in the passband. The digital passband measurement data signal SF can be supplied to an envelope 250.
The digital envelope 250 thus receives the digital passband measurement data signal SF which can reproduce a signal having positive as well as negative amplitudes. Referring to Fig. 6A, the received signal is rectified by a digital rectifier 270, and the rectified signal can be filtered by an optional low-pass filter 280 to generate a digital envelope signal SENV.
The signal SENV is thus a digital representation of an envelope signal which is generated in dependence on the filtered measurement data signal SF. According to some embodiments of the invention, the optional low-pass filter 280 can be eliminated. Such an embodiment is discussed in connection with Fig. 9 below. Thus, the optional low-pass filter 280 in envelope 250 can be eliminated when the decimator 310, which is discussed in connection with Fig. 9 below, includes a low-pass filter function.
According to the embodiment of the invention shown in Fig. 6, the signal SENV is supplied to the output 260 of the preprocessor 200. The preprocessed digital signal SMDP delivered at the output 210 (Fig. 5) is thus the digital envelope signal SENV.
Since, according to the prior art, analog devices for generating an envelope signal depending on a measurement signal use an analog rectifier whose inherent balancing error 10 is introduced into the resulting signal, the digital envelope 250 will advantageously generate a true alignment without balancing errors. The digital envelope signal SENV will thus have a good signal-to-noise ratio, since the fact that the sensor is mechanically resonant at the resonant frequency in the passband of the digital bandpass filter 240 leads to a high signal amplitude and the fact that the signal processing is performed in the digital domain eliminates addition of noise and eliminates addition of balancing errors.
Referring to Fig. 5, the preprocessed digital signal SMDP is delivered to the input 220 of the evaluator 230.
According to another embodiment, the filter 240 is a high pass filter with cut-off frequency fLC.
This embodiment simplifies the construction by replacing the bandpass filter with a high-pass filter 240, thereby leaving the low-pass filtering to another low-pass filter downstream, such as the low-pass filter 280. When the mechanical resonant frequency fRM is somewhere in the range from 30 kHz to 35 kHz, the high-pass filter 240 may be designed to have a lower cut-off frequency fLC = 30 kHz. The high-pass filtered signal is then supplied to the rectifier 270 and further to the low-pass filter 280. According to one embodiment, it should be possible to use sensors 10 which have a resonant frequency somewhere in the range from 20 kHz to 35 kHz. To achieve this, the high-pass filter 240 can be designed to have a lower cut-off frequency fLC = 20 kHz.
Fig. 6B illustrates an embodiment according to which the digital bandpass filter 240 supplies the filtered signal SF to the digital rectifier 270, and the rectifier 270 supplies the rectified signal SR directly to a state analyzer 290 (See Fig. 7 in connection with Fig. 6B).
Fig. 7 illustrates an embodiment of the evaluator 230 (see also Fig. 5). The evaluator 230 according to the embodiment of Fig. 7 comprises a state analyzer 290 which is arranged to receive a pre-processed digital signal SMDp indicating the state of the machine 6. The state analyzer 290 can be controlled to perform a selected state analysis function by means of a selector signal which is supplied on a control input 300. Examples of state analysis functions 105 are schematically illustrated by box Fig. 7. The selector signal delivered to control input 300 can be generated by user interaction via the user interface 102 (see Fig. 2A).
As mentioned above, the analysis device 14 may comprise a top level analysis function F3, 105 (See Fig. 4 & Fig. 7).
According to one embodiment of the invention, the top level analysis function may be performed by the state analyzer 290 in response to activation via the control input 300. In response to the top level analysis enable signal, the analyzer 290 activates a top level analyzer 400 (See Fig. 7), and the digital measurement signal SMDP is delivered to an input on the top-level analyzer 400.
The peak level analyzer 400 is adapted that the signal during the course of a peak monitoring time TpM in order to determine a maximum amplitude level APR which is indicative of the mechanical condition of the monitored part, i.e. layer 7 and / or axis 8. The maximum amplitude level ApR can also called representative top amplitude APR.
As mentioned above, the peak amplitude detected in the measurement signal, when the peak amplitude value originates in a mechanical vibration in the monitored machine, may be indicative of the condition of the machine. When a storage unit is monitored, the peak amplitude value may be indicative of the condition of the storage unit. In fact, the peak amplitude value may be indicative of the oil film thickness in a monitored layer. Thus, the detected peak amplitude value may be indicative of separation between the metal surfaces of the scroll interface.
The oil film thickness may depend on the supply of lubricant and / or on the alignment of the shaft. Furthermore, the oil film thickness can depend on the load on the shaft, ie on the force with which the metal surfaces are pressed against each other, the metal surfaces being, for example, a metal surface of a bearing and the metal surface of a shaft. The actual detected value of the maximum amplitude value may also depend on the mechanical condition of the bearing surfaces.
However, the ability to correctly indicate the state of the rotatable member based on a detected amplitude peak value requires that the detected amplitude peak value actually originate in the rotatable member. Machines in an industry, such as e.g. a paper mill, may be subjected to mechanical shock from tools or other machine equipment, which may cause mechanical vibrations or shock waves in the monitored machine.
Thus, an amplitude peak level in the digital measurement signal may be caused by the environment of the machine, in which case the actual highest amplitude value detected in the measurement signal may have zero and nothing to do with the state of the monitored machine part 8. Such peak amplitude levels in the measurement signal the mechanical condition of the monitored machine part 8 is considered as noise, within the scope of this document. Furthermore, electric fields in the vicinity of the sensor or in the vicinity of conductors belonging to the state measuring system can interfere and thus give rise to voltage amplitude peak values in the measuring signal.
Such voltage amplitude peaks can also be considered as noise.
The inventor realized that there is a particularly high noise level in the mechanical vibrations of certain machine equipment, and that such noise levels inhibit the detection of machine damage. Thus, conventional preventive condition monitoring methods have failed to provide a sufficiently early and / or reliable warning for future deteriorating conditions. The inventor's conclusion was that there may be a mechanical vibration VMD which indicates a deteriorated condition in such machinery, but that conventional methods for correctly detecting such vibration have hitherto been insufficient.
The inventor realized that machines having slowly rotating parts were among the types of machine equipment for which conventional methods of preventive condition monitoring have failed to provide sufficiently reliable warning of impending deterioration.
After realizing that a particularly high noise level in the mechanical vibrations of certain machine equipment inhibits the detection of machine damage, the inventor provided a method for enabling a more reliable detection of a signal amplitude peak level indicative of an incipient damage in a rotatable part 8 of the monitored machine 6.
However, tests indicated that, even in a laboratory environment where there is very little or no noise, the detected peak level often varies for a rotatable part, ie each revolution of a rotating shaft does not give rise to identical peak levels. After careful study of such amplitude levels, the inventor concluded that the amplitude levels originating from rotation of a monitored rotatable member closely follow the normal distribution, also called the Gaussian distribution; and that it is necessary to store the amplitude levels originating in a plurality of revolutions of a rotatable member in order to detect a relevant true peak value which can be used for accurately determining the state of the monitored rotatable member. In this context, it should be noted that the normal distribution is a probability distribution that describes data that accumulates around the mean. The graph of the associated probability center of gravity function is bell-shaped, with a peak at the mean, and is known as the Gaussian function or clock curve.
Figure 8 is a schematic illustration of a rectified signal SR that could be supplied by the rectifier 270 (Fig. 6B) to the peak analyzer 400 (Fig. 7). Figure 5 in connection with Fig. 6B and Fig. 7 provides an overview of an embodiment of the analysis device.
The peak level analysis F3 (See Fig. 7 & Fig. 4) is adapted to monitor the signal during the duration of a peak monitoring period TpM 1 in order to determine a relevant maximum amplitude level. The monitoring period TpM corresponds, in the example illustrated in Figure 8, 14 revolutions for the monitored rotatable part. Individual revolutions for the monitored rotatable member are indicated by the reference 405 in Fig. 8. By defining the monitoring period TpM in the form of a number of revolutions for the monitored rotatable member, rather than in the form of a fixed period of time, the quality of the analysis is thus increased. More specifically, the inventor realized that when the number of detected peak values Ap is considered in relation to the amount R of rotation of the monitored rotatable part during the measurement, statistical methods can be used to achieve improved quality of the resulting amplitude peak value.
The inventor realized that if the distribution of detected amplitude peaks A1 is similar to the Gaussian distribution, then it can be concluded that one revolution of one axis can give a different set of amplitude peaks than another rotation of the same axis.
An embodiment of the method comprises the steps of: Receiving a first digital signal due to mechanical vibrations originating in rotation of said part; Analyzing the first digital signal to detect amplitude peaks A1 during a finite time period TpM, the finite time period corresponding to a certain amount of R rotation for the rotatable part. This certain amount of R rotation should correspond to more than one revolution for the monitored rotatable part. The method further comprises defining a plurality of NR amplitude ranges RA; Sort the detected amplitude peaks A1 in the corresponding amplitude range RA to reflect the occurrence N of detected amplitude peaks AP within this plurality of amplitude ranges.
Fig. 9 illustrates a histogram resulting from a measurement in which the measurement time period TpM corresponded to fourteen (R = 14) revolutions of the monitored rotatable part under laboratory conditions without any noise, i.e. each of the black dots corresponds to a detected amplitude. -peak value Ap. Thus, the value of “certain amount R rotation” is R = 14.0 revolutions, and the fi rivet time period TpM was the time it took for the monitored part 8 to rotate 14 revolutions. The monitored part 8 can be a shaft. Thus, according to embodiments of the invention, the measurement time period TpM may depend on the rotational speed of the rotatable member so that when the monitored rotatable member rotates at a slower speed, the measurement time period TpM becomes longer, and when the monitored rotatable member rotates at a higher speed, the measurement time period TPM becomes shorter.
Based on the knowledge that the measurement was performed during R = 14 complete turns of rotation for the monitored rotatable part, and assuming that a maximum amplitude peak value is detected once per turn, it can be seen in Figure 9 that the highest fourteen (14 The detected amplitude values actually vary slightly, the highest amplitude being indicated by the reference 410 and the 14th highest amplitude being indicated by the reference 420. Thus, it can be deduced from Figure 9 that the amplitude peak value Ap detected during a revolution often differs. from the amplitude peak value detected during another revolution. In other words, if measurement were to be performed during a single revolution, then a number of one-revolution measurements on the nursing shaft would result in quite large variations in the detected amplitude peak value.
The inventor realized that it is desirable to provide a measurement method which is reliable in the sense that it should give repeatable results. When the measurement procedure is repeatedly performed on the same rotatable part so that a plurality of monitoring periods TPW, TpMg, TpMg, TPM4, TpM5 result in measurement results in the form of a plurality of representative amplitude peaks Apm, APPZ, APP3, APP4, generated in immediate time sequence, then it is desirable that these fl ertal representative amplitude peak values APPT, APP2, APPS, APP4 have substantially the same value.
By performing a large number of test measurements in a laboratory environment where there was very little or no noise, the inventor concluded that it is desirable to monitor a rotating part during a finite time period TPM corresponding to several revolutions R to detect a true amplitude peak APT which is indicative of the mechanical condition of the monitored part, i.e. bearing 7 and / or shaft 8. In this context, the true amplitude peak value APT is true in the sense that it actually derives from a mechanical vibration VMD caused by relative movement between metal surfaces in a monitored part, such as a bearing ball and an inner ring surface, and is not caused by any noise or disturbance.
The choice of value of the parameter R is an issue that requires careful balancing, as monitoring during a single revolution, ie R = 1, would probably lead to an excessively low amplitude peak value APT which may therefore be insufficient to indicate the mechanical condition. for the monitored rotating part. If, on the other hand, the rotating part is monitored for an extremely long time, which approaches infinity in statistical terms, then the detected amplitude peak value APT will slowly increase towards the infinite, which in reality means that after an extremely long time of operation a rotating part associated with a bearing unit will break. Thus, the inventor concluded that it is necessary to find a balanced value for the parameter R, so that on the one hand one has an R-value which is high enough to detect a true amplitude peak value APT which is indicative of the mechanical condition of the monitored part. , while on the other hand you have an R-value that is low enough to keep the measurement period TPM to a reasonable finite duration.
Based on a large number of test measurements under substantially noise-free conditions, the inventor could conclude that it is desirable to monitor a rotating part for a period of time TPM corresponding to a certain amount of rotation R of the rotating part; wherein the determined amount of rotation R corresponds to at least eight (R = 8) revolutions of the monitored rotatable member to actually detect a true amplitude peak value APT indicative of the mechanical condition of the monitored member. Based on these test measurements, the inventor concluded that monitoring the rotating part 10 for a finite time period TPM corresponding to at least ten (R = 10) revolutions of the monitored rotatable part gives a more accurate true amplitude peak value APT, i.e. a true amplitude peak value APT which is more accurately indicative of the mechanical condition of the monitored part. This conclusion is based on experiments indicating that a further increase of the finite time period TPM to a finite time period of more than ten (R = 10) revolutions, in an environment free from noise, can lead to a detection of a higher true amplitude peak APT, but the increase in the detected true amplitude peak APT is small relative to the increased monitoring time period TPM.
When measuring and collecting amplitude peak values AP is performed for a period of time TPM corresponding to R = 14 complete revolutions of the monitored rotatable part, and then the amplitude peak values AP are organized in a histogram, as illustrated in Figure 9, the amplitude the peak values AP which are sorted into the amplitude level 420 for the 14th highest detected amplitude very stable. It can be seen from the histogrammetry of Figure 9 that the four amplitude peak values were detected at that amplitude range 420. Thus, a stable measurement value can be achieved, i.e. an achievement of substantially the same amplitude peak value when performing a plurality of measurements on the same rotating part, by to focus on the Rth highest amplitude value, where R is a number indicating the number of revolutions performed by the monitored rotatable member during the peak level monitoring time TPM.
Fig. 10 illustrates a histogram resulting from a measurement where the peak level monitoring time TPM corresponded to fourteen (14) revolutions of the monitored rotatable member. The histogram of Figure 10 is the result of an experiment in which two very high amplitude mechanical disturbances 430, 440 were generated during the peak level monitoring time TPM. The two signal peaks corresponding to the two very high amplitude mechanical disturbances 430, 440 are also illustrated in Figure 8.
Experience and a number of measurements have indicated that when monitoring a machine that has a part that rotates at a rotational speed, the highest peak amplitude value, originating in an incipient damage, is a very relevant amplitude value in preventive maintenance.
However, since the highest peak amplitude value does not occur each time the monitored shaft rotates an entire revolution, it is necessary to monitor a rotatable member for a period of time that allows multiple revolutions. Unfortunately, however, a longer measurement time in a real situation means increased noise level in the measurement signal. In an industrial environment, such as a paper mill, other machine equipment in the vicinity of the monitored machine can occasionally cause mechanical vibrations or shock pulses, and the longer the measurement time, the greater the risk that such external mechanical vibrations will cause the highest detected amplitude peaks. For these reasons, the measurement method, which aims to provide a reliable and repeatable achievable representative peak value, must meet the conflicting requirements of: On the one hand, involving measurement for a sufficiently long time to collect amplitude peaks over a plurality of revolutions. for the monitored rotating part so that an amplitude peak value representative of the highest amplitude peak value caused by the state of the monitored rotating part is collected, and on the other hand avoid that the measuring method requires so long a time that noise caused by e.g. machinery in an industrial environment corrupts the measurement results.
According to one embodiment of the invention, the Rzte highest amplitude is selected to be a representative amplitude peak APR.
However, the Gaussian function or clock curve is by nature such that the frequency of low amplitude values actually tells us something about the amplitude of the not-so-common-highest peak amplitude peak values.
According to one aspect of the invention, the invention comprises estimating a representative amplitude peak value (APR) in dependence on the sorted amplitude peak values (AP) and said determined amount (R).
According to one embodiment, the estimating step comprises the creation of an accumulated histogram.
Fig. 11A is a flow chart illustrating an embodiment of a method of operating the device 14 so that it is prepared to perform a peak value state analysis. The procedure of Fig. 11A can be performed when an embodiment of the analysis function F3 (See Fig. 4 & Fig. 7) is run on the processor 50 (See Fig. 2A). In a step S10, a parameter value R is set, and in another, optional step, S20, a parameter n can be set. According to one embodiment, the parameters R and n, respectively, can be set in connection with the manufacture, or in connection with deliveries of the measuring device 14.
Thus, the parameter values R and n can be preset by the manufacturer of the apparatus 14, and these values can be stored in the non-volatile memory 60 (see Fig. 2A).
Alternatively, the parameter values R and n can be set by the user of the apparatus 14 before a measurement session is performed.
Fig. 11B is a flow chart illustrating an embodiment of a method of operating the device 14 so that it performs peak value state analysis. The method according to Fig. 11B is can be performed when an embodiment of the analysis function F3 (See Fig. 4 & Fig. 7) is run on the processor 50 (See Fig. 2A). In a step S50, a current speed value fROT is read, and stored in a data memory 52. When the part 8 to be monitored rotates at a constant rotational speed, the speed value fROT can be entered by a user via the user interface 102 (Figure 2A). When the rotational speed fROT of the monitored part is variable, a speed detector 450 (See Figure 1 & Figure 5) can be provided to deliver a signal indicating the rotational speed fROT of the shaft 8. The rotational speed of the shaft 8 fROT can be provided in as number of revolutions per second, rps, ie. Hertz (Hz) to an input 460 of the digital signal processing means 180 (see Figure 5) so that one can be used by the processor 50 as it runs the program to perform the amplitude peak analysis function.
In step S60 further preparations for the measurement session step S70. The preparations according to step S60 may include preparing a suitable table 470 for the data to be collected.
Fig. 13B is a schematic illustration of a plurality of memory locations arranged as a table 470, and suitable for storing data to be collected. Table 470 may be stored in memory 52 (Fig. 2A) or a memory internal to processor 50.
Fig. 13A illustrates a histogram with a plurality of amplitude compartments 500, indicated individually by the references r1 to r750, each amplitude compartment r1..r750 representing an amplitude level A1. Although Figure 13 shows 750 (seven hundred and fifty) amplitude compartments, it is only a sample number. The number of amplitude compartments can be set to a suitable number in step S60 (See Fig. 11B) by the user, via the user interface 102 (Figure 2A). Figure 13A is comparable to Figure 10, as both figures show a number of amplitude compartments along one axis 480, and the presence of detected amplitude peaks along another axis 490. The amplitude axis 480 may have a certain resolution, which can also be set by the user, via the user interface 102. Alternatively, the resolution of the amplitude axis 480 can be preset.
According to one embodiment, the resolution of the amplitude axis 480 can be set to 0.2 dB, and the amplitudes to be recorded can range from a minimum amplitude of Ar1 = -50 dB to a maximum amplitude of A fl50 = +100 dB.
Referring to Fig. 13B, the illustrated table is a representation of the histogram shown in Fig. 13A having amplitude compartment 500, indicated individually by the references r1 to r750, each amplitude compartment r1..r750 representing an amplitude level A ,. Table 470 also includes memory positions 510 for amplitude values A 1, and memory positions 520 for variables N, which reflect the occurrence.
The compartment r1 is associated with an amplitude value An and with a memory position 520 for a variable Nm for storing a value indicating how many times the amplitude A, 1 has been detected.
In step S60 (Figure 11B), before starting the measurement session S70, all the occurrence variables N, 1 to N fl50 can be set to zero (0). Then the measurement session S70 can begin.
The measuring session S70 may comprise receiving a first digital signal SR, SMDP due to mechanical vibrations originating in rotation of said part (See Fig. 6B &7); and Analyzing the first digital signal SR, SMDP to detect amplitude peaks A1 during a finite time period Tpm, the finite time period corresponding to a determined amount R of rotation of the rotatable member 8; wherein the determined amount of R rotation corresponding to more is one revolution of rotation of the monitored rotatable member; and sorting the detected amplitude peaks AP into the corresponding amplitude range 500 to reflect the occurrence N of detected amplitude peaks Aβ within said plurality of ranges 500 (See Figure 13B).
The duration of the measuring session is controlled depending on the amount of rotation of the rotatable part, so that the rotatable part rotates at least R revolutions, as mentioned above. Step S80 in Figure 11B represents the step of controlling the duration of the finite time period Tpm in that way. A tachometer may be provided to monitor the signal fROT so as to ensure that the measurement session continues for the duration of the fi nine time period Tpm, corresponding to the determined amount R of rotation of the rotatable member 8. Alternatively, the detector 450 may generate a signal indicating the amount of rotation, and the duration of the measurement can be controlled only in the dependent amount of rotation of the rotatable part 8, independent of time. Alternatively, the duration of the measurement session Tpm can be controlled depending on time information provided by the clock 190 (Fig. 5) in conjunction with the rotation speed information fROT provided by the detector 450 so that the duration Tpm is adjusted to ensure monitoring is performed during the desired amount of rotation n * R. In this context, it is noted that R is a positive number greater than one, and n is a positive number equal to one (1) or greater than one (1). The parameter R can be an integer, but can alternatively be a decimal number. The parameter n can be an integer, but can alternatively be a decimal number. In the example shown in Figure 8 above, the parameter R = 14 and the parameter n: 1.
In step S90 (Figure 11B), a representative amplitude peak value APR is determined in dependence on the sorted amplitude peak values Aβ collected during the measurement session S70.
F ig. 12A is a flow chart illustrating an embodiment of a method for performing step S70 so that the peak measurement session is performed.
In a step S100 a digital signal SR, SMDp is received which is due to mechanical vibrations of the top level analyzer 400 (See Fig. 7). When a signal peak is detected (step S110), the amplitude peak value of the detected peak is measured (step S120), and a corresponding amplitude interval r .
In a step S140, the corresponding instance count value N fi is incremented by one unit to represent the detection of a peak within the amplitude interval compartment n.
Then, step S80 in Figure 11B is performed to determine whether the measurement session is complete or should continue. If it is to continue, steps S100 to S140 must be repeated, ie. step S70 in Figure 11 is performed again. When step S80 determines that the measurement session is complete, a representative amplitude peak value APR is determined (S90) depending on the sorted amplitude peak values Aβ collected during the measurement session S70, as mentioned above.
Since the measurement results illustrated in Figure 9 represent a maximum peak amplitude 410 detected during R = 14 revolutions under substantially noise-free conditions, the highest peak 430 was generated, in the measurement session illustrated in Figure 10 during R = 14 revolutions, depending on a interference, i.e. it reflects noise, and as such the top 430 does not carry any information about the condition of the rotating part 8. Thus, it is desirable to obtain a representative amplitude peak APR based on signal values that reflect measured values provided by the sensor 10 in response to vibrations originating from the shaft and / or the bearing as the shaft rotates. Especially when measuring slow rotating parts, which thus require a longer measuring period TpM when measuring to be performed over a certain predetermined amount of rotation R, the amount of noise can also increase due to the longer measuring session required due to the lower rotational speed . Thus, there is a need for a robust measurement method that can withstand noise. In a wind turbine application, the shaft whose bearings are to be analyzed can rotate at a speed of less than 120 revolutions per minute, ie the shaft rotation frequency fROT is less than 2 revolutions per second (rps). Sometimes such a shaft-to-rotate is rotated at a speed of less than 50 revolutions per minute (rpm), i.e. the shaft rotation frequency fRQT is less than 0.83 rps. In fact, the rotational speed can typically be lower than 15 rpm. Since an axis with a rotational speed of 1715 rpm, as discussed in the above book, produces 500 revolutions in just 17.5 seconds; then it takes ten minutes for a shaft rotating at 5 revolutions per minute to achieve 500 revolutions. Some large wind turbines have shafts that typically rotate at 12 rpm = 0.2 rps. At 12 rpm, it is more than four minutes to achieve fifty full revolutions, and thus the risk of impact noise occurring during the measurement is much higher when the peak level analysis is to be performed on a rotating part that has such a low rotational speed. Similarly, some machine parts in paper mills rotate at a speed of less than 50 rpm.
Figure 14A is a flow chart illustrating an embodiment of a method for determining a representative amplitude peak value APR based on the amplitude peak values Aβ collected during the measurement session S70. The method according to the embodiments of Figure 14A illustrates a method by which high-amplitude noise can be sorted out. Thus, the method of Figure 14A can be advantageously used for top-level anays of rotatable parts having a speed of less than 50 rpm.
In a step S150, data relevant to the analysis is read. This includes the value of parameter R, which is used during measurement session S70, and the value of parameter n. It may also include peak value measurement data in histogram format, as illustrated in Figs. 13A, 13B or 13C. The peak value measurement data to be analyzed may be the information collected as described above, for example in connection with steps S70 & S80 above and / or as described in connection with Figs. 12A or 12B.
In step S160, identify the nth highest detected amplitude peak value. Referring to Figure 13B, and assuming that the data is sorted so that the highest amplitude compartment is on the right side of the table according to Figure 13B (ie amplitude A fl50, associated with the compartment r750, represents the highest detectable amplitude value), starting with occurrence N fl50, move to the left and add occurrence values N fl until the sum is equal to n. Once the nth highest detected amplitude has been found, the subsequent step S170 includes identifying the amplitude compartment r, which represents the nth highest detected the amplitude peak value and the corresponding amplitude value Ari.
In the subsequent step S180, select the identified amplitude value Afi to be an estimate of the representative amplitude peak value APR: ÅPRI: Year: According to one embodiment, with reference to steps S10 and S20 in Figure 11A, the parameter R is set to 10, and the parameter n is set to 5, which leads to the measurement and collection of peak amplitude values during n * R = 5 * 10 = 50 revolutions of the monitored part (steps S70 and S80 in Figure 11B).
If a true peak value is generated at least once during R turns, and there is also some high-amplitude noise in the form of false peak values, then according to this embodiment the four highest peak values can be sorted out and the method will still identify a true peak value in in the form of the nth highest detected peak value, ie the fifth highest detected peak value. Thus, assuming that the amount of high amplitude interference causes at most four of the highest five peak values, this embodiment delivers the amplitude of the nth peak value as a representative amplitude peak value APR.
Table 1 below illustrates some examples of combinations of parameter settings for R and n, together with the resulting duration of the measurement session corresponding to the ability to sort out noise.
Number of removed- The measuring session's sorted R n tity (lap) noise peaks 10 6 60 5 10 7 70 6 10 8 80 7 10 9 90 8 10 1 0 100 9 10 1 1 1 10 10 10 1 2 120 1 1 10 1 3 130 12 10 14 140 13 9 6 54 5 9 7 63 6 9 8 72 7 9 9 81 8 9 10 90 9 9 1 1 99 10 9 12 108 1 1 Table 1: However, the inventor also concluded that since the distribution of true amplitudes tud peak values originating in rotation of a monitored rotatable part closely follow the normal distribution, so it may be possible to estimate an amplitude peak value which statistically rarely occurs on the basis of detected amplitude peak values which occur more frequently. Based on this insight, the inventor continues to develop a further advantageous method of estimating a representative amplitude peak value APR depending on the sorted amplitude peak values Aβ and the amount of rotation R of the monitored part, as discussed below in connection with Figure 14B.
Fig. 14B is a fate diagram illustrating a further embodiment of a method for estimating a representative amplitude peak value APR on the basis of the amplitude peak values Aβ collected during the measurement session S70. The method of Fig. 14B may be an embodiment of step S90 of Fig. 11B.
In a step S200, a parameter g is set to a value (n * R) / q1. 9 1 = (f1 * R) / q1 The parameter q1 can have a numerical value 1 or greater than 1. According to embodiments of the invention, the parameter q1 is preset to a value between one (1) and three (3).
In a step S210, an amplitude interval rg is identified (See Figure 13) which holds the gth highest detected amplitude peak value.
In a step S220, a parameter h is set to a value (n * R) / q2. h: = (n * R) / q2 According to embodiments of the invention, the parameter q2 is preset to a value between one (2) and three (5). The value of the parameter q; is always greater than the value of the parameter q1.
Q2> Q1 In a step S230 an amplitude interval rf is identified (See Figure 13) which holds the highest highest detected amplitude peak value.
In a step S240, an estimate of a representative amplitude peak value APR is obtained based on the values (rg, g) and (rh, h). This will be explained in more detail below in connection with Figure 15A.
Setting the parameters n = 5, R = 10 and q1 = 1 in step S200 Before g = 50. Thus, the measurement session comprises 50 revolutions (since n * R = 50), and setting g = 50 indicates that we identify the position in the histogram where the 50th highest detected pulse is stored. Thus, we identify the position of the histogram according to Figure 13 where pulses that occur at a frequency of once giving revolutions will be reflected.
Similarly, setting the parameter q2 = 4 in step S200 leads to h = n * R / q2 10 15 20 25 30 37 = 12.5. Thus, the measurement session comprises 50 revolutions (since n * R = 50), and setting h = 12 indicates that we identify the position in the histogram where the 12th highest detected pulse is stored. Thus, we identify the position of the histogram according to Figure 13 where pulses that occur with a frequency of one turn every four turns will be reflected.
Since the distribution of true amplitude peak values originating in rotation of a monitored rotatable part follows close to the normal distribution, these two positioning histograms can be used to estimate an amplitude peak value that statistically occurs less frequently. As mentioned above, the representative amplitude peak value ApR can be an amplitude that statistically occurs once every ten revolutions. Thus, when setting the parameter R to the value 10, the method involves estimating the amplitude of a peak value that occurs once in ten revolutions, based on observation of the occurrence frequency and amplitude of peaks that occur once per revolution and once in four revolutions.
Advantageously, this method enables the sorting out of 11 false high amplitude peak values, since it still allows the estimation of an accurately representative amplitude peak value APR, when the parameters g and h, respectively, are set as mentioned above. Furthermore, it should be noted that this method enables the removal of 11 false high amplitude peaks as it reduces the required duration TPM for the measurement session to the duration of only 50 revolutions. This is because n * R = 5 * 10 = 50. This effect is advantageously achieved because the parameters q1 and q2 are selected so that the two parameters g & h are selected to values representing relatively high frequency frequencies of amplitude peaks, and the amplitudes of the high frequency peaks are used to estimate an amplitude peak value which statistically occurs less frequently.
Thus, the embodiment of Figure 14B enables substantially the same accuracy in estimating representative amplitude peak value APR based on measurement under 50 revolutions as the method of the embodiment of Figure 14A provides based on measurement under 120 revolutions (compare Table 1 above).
According to an embodiment of the invention, the estimation can be performed by generating an accumulated histogram table which represents all amplitudes detected in a measurement session and their occurrence frequency. Fig. 13C is an illustration of a cumulative histogram table corresponding to the histogram table of Fig. 13B. The cumulative histogram table of Fig. 13C includes the same number of amplitude interval bays as the table of Fig. 13B. In the cumulative histogram, the occurrence N 'is represented as the number of occurrences of detected peaks having an amplitude exceeding the amplitude A,' according to the associated amplitude compartment r. This advantageously provides a smoother curve when the cumulative histogram is plotted. Since a "normal" histogram that reproduces a limited number of observations will reproduce a lack of observations N at an amplitude compartment such as a notch or 'dent' at that compartment, the cumulative histogram will provide a smoother curve, making it more suitable for use in estimating occurrence at one amplitude level based on observation of occurrences at other amplitude levels.
According to an embodiment of the invention, the amplitude levels are represented as logarithmic values, and also the cumulative occurrence is represented by the logarithmic value of the cumulative occurrence.
Figure 15A is an illustration reflecting the principle of a cumulative histogram resulting from a measurement, corresponding to the table of Figure 13C. Although a cumulative histogram using real detected values may take a different form from that shown in Figure 15A, Figure 15A illustrates the principle of estimating the representative amplitude peak value APR which reflects the amplitude level occurring every Rzte revolution.
One axis 542 of the cumulative histogram reflects occurrence, and the other axis 544 reflects amplitude. When n = 5, R = 10 and q1 = 1, g = 50 which represents 50 occurrences, which also corresponds to one occurrence per revolution. One occurrence per revolution can be written "1/1". Thus, the axis 542, in the cumulative histogram reflecting before occurrence, can represent g as "1/1 Similarly, h can represent an occurrence of one per four, also expressed as" 1 / 4 ", and when R = 10, R can reflect an occurrence of one in ten, also expressed as" 1/10 "(see Figure 15A).
The parameter values rg, g and n., H can be determined in the manner described above in connection with Figure 14B. The parameter values rg, g reflect a point 550 in the cumulative histogram indicating peaks that occur once per revolution. The parameter values rh, h reflect a point 560 in the cumulative histogram indicating peaks that occur once every four turns. The inventor realized that the logarithmic cumulative histogram in this part of the normal distribution curve has good agreement with a straight line, which makes it possible to draw a straight line 570 through points 550 and 560. When that line 570 is extended it will cross a line 580 representing the R-occurrence at a point 590. The amplitude value for point 590 represents the amplitude level APR that occurs every Rzte revolution.
On the basis of experiments, the inventor determined that the parameter q1 would advantageously have a value not lower than one (1), since selection of the parameter q1 to a value less than one can lead to poor results of the estimation procedure because a cumulative history grams which represents a bearing unit which has an outer ring damage deviates comparatively much from a straight line, thus leading to a larger error in the estimation.
Noise echo suppression Furthermore, the inventor realized that impact noise in industrial environments, which can be caused by an object hitting the casing of the machine having a monitored rotating part 8, can cause shock waves traveling back and forth, echoing in the machine body. Thus, such echoing shock pulses can be captured by the sensor 10 (Figures 1, 2A, 5) and reproduced in the resulting signal SR, SMDP (Figs. 6B, 7) as a burst of amplitude peaks.
Thus, such a burst of amplitude peaks can unfortunately ruin a peak-level analysis, unless the consequences of such bursts can be reduced or eliminated.
Fig. 12B is a flow chart illustrating an embodiment of a method for performing step S70 (Figure 11B) for performing the peak value measurement session and further addressing the influence of bursts of noise amplitude peaks.
Step S300 in the method embodiment illustrated in Figure 12B may be performed after step S60, as described in connection with Figure 11B above. In step S300 the reader peak level analyzer of the current rotational speed fROT, which can be supplied from the speed detector 450, as described above (See Figure 5). Reading a real-time value of the rotational speed fRQT advantageously makes it possible for this method to be performed even when the rotatable part to be analyzed rotates at a variable rotational speed. In a step S310, an echo cancellation period Tes is calculated. The eco-suppression period Tes is set to: Tes = 1 / (e ~ fROT) Where e is a factor that has a value equal to ten or less than ten: e <= 10 An effect of the eco-suppression procedure is to reduce the number of peak values per revolutions for the monitored part 8 to a maximum of e peaks per revolution. Choosing e = 1O thus leads to the delivery of a maximum of 10 peaks per round. In other words, the echo suppression period Tes will have a duration corresponding to the duration of one tenth of a revolution when e = 10.
In a step S320 the measurement signal SMDP, SR to be analyzed is received, and in a step S330 the amplitude of the received signal SR is analyzed to detect any received peak values. In a step S340, detected peak values Aβ are supplied with a frequency of fes or less, each delivered amplitude peak value reflecting the highest detected amplitude during the echo cancellation period Tas. This is done so that it may be longer than an echo cancellation period Tes between two consecutively delivered output values from the echo canceller, but the period between two consecutively delivered output values from the echo canceller will never be shorter than the echo cancellation period Tes.
In a subsequent step S350, the peak values Ap supplied by the echo canceller are received by a log generator. The log generator calculates the logarithm of the peak value Ap in real time.
In a step S360 the amplitude compartment corresponding to the relevant peak value Ap is identified in a histogram table 470 and / or 530 (See histogram table 470 and cumulative histogram table 530 in Figures 13B and 13C, respectively), and in a step S370 the corresponding occurrence counter value N fi, N fi 'is increased. up with a device.

权利要求:
Claims (29)
[1]
A method of operating an apparatus for analyzing the state of a machine having a rotating speed (fPOT) member, comprising the steps of: receiving a first digital signal (SMD, SP, SP) due to mechanical vibration ions originating in rotation of said part; analyzing the first digital signal to detect amplitude peaks (Ap) during a finite time period (TPm), the finite time period corresponding to a certain amount (R) of rotation of the rotatable member; wherein the determined amount (R) of rotation corresponds to more than one revolution of the monitored rotatable member; define a plurality (NP) amplitude range; sort the detected amplitude peaks (Aβ) into corresponding amplitude intervals so that the occurrence (N) of detected amplitude peaks (Aβ) within said plurality of amplitude ranges is represented; estimate a representative amplitude peak value (APR) depending on the sorted amplitude peak values (Ap) and the determined amount (R).
[2]
The method of claim 1, further comprising delivering the representative amplitude peak value (APR) to a user interface for presentation to a user.
[3]
The method of claim 1 or 2, further comprising performing a state analysis function (F1, F2, Fn) to analyze the state of the machine depending on the representative amplitude peak value (APR).
[4]
The method of any preceding claim, wherein the estimating comprises selecting the Rth highest amplitude to be the representative amplitude peak value (APR).
[5]
The method of any of the preceding claims, wherein the estimating comprises creating an accumulated histogram.
[6]
The method according to any one of the preceding claims, wherein the amplitude levels originating in rotation of the monitored rotatable part correspond well with the normal distribution, also called the Gaussian distribution; and wherein amplitude levels originating in a plurality of turns of the rotatable member are stored to detect a relevant true peak value used to determine the state of the monitored rotatable member.
[7]
The method of any of the preceding claims, wherein the estimating step comprises estimating a not-so-frequent highest amplitude peak value (APR, 590) based on the Gaussian function or clock curve being such that an occurrence frequency of low amplitude values ( 550, 560) provides information on the amplitude of not-so-frequent highest peak amplitude values (APR, 590).
[8]
The method of any of the preceding claims, wherein the determined amount of rotation comprises at least n ~ R turns, wherein n is a number having a numerical value of at least one and R has a numerical value of at least 8.
[9]
The method of any of the preceding claims, wherein the numerical value of n is at least 2; and the estimating step comprises selecting the nth highest detected peak amplitude.
[10]
The method of any of the preceding claims, wherein the numerical value of R is at least 10.
[11]
The method of claim 7, 8 or 9, wherein the step of estimating a not-so-frequent highest amplitude peak value (APR, 590) uses detected amplitude peaks (Ap, rg, n., 550) having an average frequency of occurrence. if one per revolution (g), and / or less than one per revolution (h), to estimate an amplitude value (APR, 590) for a peak having an average frequency of occurrence of one per eight revolutions or less than one per eight turns (R); wherein the detected amplitude peak values (Ap, rg, rh, 550) used for the estimation step have an average frequency of occurrence that is higher than one per eight revolutions (R). 10 15 20 25 30 43
[12]
The method of claim 11, wherein the detected amplitude peaks (Ap, rg, n., 550) used for the estimation step have an average frequency of occurrence greater than one per five revolutions.
[13]
The method according to any of the preceding claims, wherein the analysis of the first digital signal comprises a burst removal step (S330, S340) which functions to supply detected peak values (Ap) with a delivery frequency fes, wherein fes = ei »fROT, where fROT is the rotational speed, and e is a factor that has a value of ten or less than ten.
[14]
The method of claim 13, wherein the burst removal step (S330, S340) operates so that each amplitude peak value (Ap, rg, n., 550) it delivers represents the highest amplitude value detected within the immediately preceding echo. the suppression period (Teg), where the echo suppression period (Tas) is the inverse of the delivery frequency phase.
[15]
An apparatus for analyzing the state of a machine having a part rotating at a rotational speed (fROT), comprising: means for receiving a first digital signal (SRED, SMD, SENV) due to mechanical vibrations originating in rotation of said part; means for analyzing the first digital signal (SRED, SMD, SENV) for detecting amplitude peaks (Ap) during a finite time period (Pm), the finite time period corresponding to a certain amount (R) of rotation of the rotatable part ; wherein the determined amount of rotation (R) corresponds to more than one revolution of the monitored rotatable part; means for defining a plurality of (NR) amplitude ranges; means for sorting the detected amplitude peaks (Ap) into corresponding amplitude ranges so as to represent the presence (N) of detected amplitude peaks (Ap) within said plurality of amplitude ranges; means for estimating a representative amplitude peak value (APR) depending on the sorted amplitude peak values (Ap) and the determined amount (R). 10 15 20 25 30 44
[16]
The device of claim 1, further comprising a user interface; and means for delivering the representative amplitude peak value (APR) to the user interface for presentation to a user.
[17]
The apparatus of claim 15 or 16, further comprising means for performing a state analysis function (F1, F2, Fn) for analyzing the state of the machine depending on the representative amplitude peak value (APR).
[18]
The device according to any of the preceding claims, wherein the means for estimating comprises means for selecting the Rth highest amplitude to be the representative amplitude peak value (APR).
[19]
The device according to any of the preceding claims, wherein the means for estimating comprises means for creating an accumulated histogram.
[20]
The device according to any one of the preceding claims, wherein the amplitude levels originating in rotation of the monitored rotatable member correspond well to the normal distribution; and further comprising means for detecting amplitude levels originating in a plurality of turns of the rotatable member to detect a relevant true peak value indicative of the state of the monitored rotatable member.
[21]
The apparatus of any preceding claim, wherein the means for estimating comprises means for estimating a not-so-frequent highest amplitude peak value (APR, 590) based on the Gaussian function or clock curve being such that an occurrence frequency of low amplitude values (550, 560) provide information about the amplitude of not-so-frequent highest peak amplitude values (APR, 590)
[22]
The apparatus according to any one of the preceding claims, wherein the determined amount of rotation comprises at least n «R turns, wherein n is a number having a numerical value of at least one and R has a numerical value of at least 8. 10 15 20 25 30 45
[23]
The device of claim 22, wherein the numerical value of n is at least 2; and the estimating means comprises means for selecting the nth highest detected peak amplitude.
[24]
The device of claim 22 or 23, wherein the numerical value of R is at least 10.
[25]
The device according to claim 21, 22 or 23, wherein the means for estimating a not-so-frequent highest amplitude peak value (APR, 590) is arranged to use detected amplitude peak values (AP, rg, rh, 550) which have an average occurrence frequency of one per revolution (g), and / or less than one per revolution (h), to estimate an amplitude value (APR, 590) for a peak having an average frequency of occurrence of one per eight revolutions or less than one per eight turns (R); wherein the detected amplitude peak values (Ap, rg, rg, 550) used for the estimation step have an average occurrence frequency that is higher than one per eight revolutions.
[26]
The device of claim 25, wherein the detected amplitude peaks (Aβ, rg, rh, 550) used for the estimation step have an average frequency of occurrence greater than one per five revolutions.
[27]
The apparatus according to any one of the preceding claims, wherein the analyzing means for analyzing the first digital signal comprises a burst remover arranged to supply detected peak values (Ap) with a delivery frequency fgs, wherein feg = e ~ fROT, where fROT is the rotational speed, and e is a factor having a value of ten or less than ten.
[28]
The apparatus of claim 27, wherein the burst remover is arranged to deliver each amplitude peak value so that each delivered amplitude peak value represents the highest amplitude value detected during the immediately preceding echo cancellation period (Tas), the echo cancellation period (Tas) being the inverse of the delivery frequency fes.
[29]
The apparatus for analyzing the state of a machine having a portion rotating at a rotational speed (fRO-f), further comprising: an input (42) for receiving an analog measurement signal (SEA) indicative of a vibration signal signature having a vibration frequency (f SEA); an A / D converter (40, 44) for generating a digital measurement signal (SMD) which depends on the analog measurement signal (SEA), the digital measurement signal (SMD) having a first sampling frequency (fs), the first sampling frequency ( fs) is at least twice (k) as high as the vibration frequency (f SEA).

类似技术:

---

公开号 | 公开日 | 专利标题

---

SE1000631A1|2011-07-19|A state monitoring system

---

US11054301B2|2021-07-06|Method and a system for analysing the condition of a rotating machine part

---

US11015972B2|2021-05-25|Apparatus for monitoring the condition of a machine

---

DK2810027T3|2017-08-28|DEVICE AND PROCEDURE FOR ANALYZING THE STATE OF A MACHINE WITH A ROTATING PART

---

DK2370801T3|2018-01-22|Analysesystem

---

AU2009330744B2|2015-04-09|Method and apparatus for analysing the condition of a machine having a rotating part

---

US9964430B2|2018-05-08|Apparatus and a method for analyzing the vibration of a machine having a rotating part

---

US20110285532A1|2011-11-24| analysis system

---

AU2015203801B2|2016-09-22|Method and apparatus for analysing the condition of a machine having a rotating part

---

AU2015203361B2|2017-06-29|An apparatus and a method for analysing the vibration of a machine having a rotating part

同族专利:

---

公开号 | 公开日

---

US10330523B2|2019-06-25|

---

EA024339B1|2016-09-30|

---

EA201290660A1|2013-01-30|

---

US20190391004A1|2019-12-26|

---

US9279715B2|2016-03-08|

---

US20120296582A1|2012-11-22|

---

SE535559C2|2012-09-25|

---

EP2526389A1|2012-11-28|

---

EP2526389A4|2017-11-15|

---

US20160290854A1|2016-10-06|

---

WO2011087440A1|2011-07-21|

---

CN102822644B|2016-02-03|

---

CN102822644A|2012-12-12|

引用文献:

---

公开号 | 申请日 | 公开日 | 申请人 | 专利标题

---

---

US3554012A|1968-02-29|1971-01-12|Iko Konsultbyron Ab|Method and arrangement for determining the mechanical state of machines|

---

US4528852A|1982-10-21|1985-07-16|Spm Instruments U.S. Inc.|Method and instrument for determining the condition of an operating bearing|

---

JPS60195426A|1984-03-16|1985-10-03|Komatsu Ltd|Fault diagnosing method of rotary mechanism|

---

DE3424692A1|1984-07-05|1986-02-06|Kletek Controllsysteme GmbH & Co KG, 2820 Bremen|Method and measuring arrangement for analysing periodic or quasi-periodic signals, in particular sound signals in the case of machines and systems|

---

GB2190198A|1986-04-28|1987-11-11|Vipac Pty Ltd|Vibrational analysis system for a machine|

---

US5258923A|1987-07-22|1993-11-02|General Electric Company|System and method for detecting the occurrence, location and depth of cracks in turbine-generator rotors|

---

JPH01127934A|1987-11-12|1989-05-19|Hitachi Ltd|Damage diagnosing device|

---

US4912661A|1987-12-01|1990-03-27|Hewlett-Packard Company|Tracking and resampling method and apparatus for monitoring the performance of rotating machines|

---

JPH01178814A|1988-01-08|1989-07-17|Mitsubishi Heavy Ind Ltd|Apparatus for confirming structure on whole based on signal|

---

GB2228088B|1988-12-16|1992-09-16|Nippon Seiko Kk|Method and apparatus for detecting cracks in bearings|

---

JPH04279826A|1991-03-08|1992-10-05|Nkk Corp|Abnormality diagnostic method and device for adjustable speed rotating system|

---

US5201292A|1991-08-30|1993-04-13|Loral Aerospace Corp.|Apparatus and method for detecting vibration patterns|

---

US5501105A|1991-10-02|1996-03-26|Monitoring Technology Corp.|Digital signal processing of encoder signals to detect resonances in rotating machines|

---

US5365787A|1991-10-02|1994-11-22|Monitoring Technology Corp.|Noninvasive method and apparatus for determining resonance information for rotating machinery components and for anticipating component failure from changes therein|

---

US5445028A|1992-09-18|1995-08-29|Ametek Aerospace Products Inc.|Dynamic digital tracking filter|

---

RU2171007C2|1993-11-09|2001-07-20|Моторола, Инк.|Level detector and method of detection of level of input signal|

---

US5633811A|1994-12-09|1997-05-27|Computational Systems, Inc.|Hand held data collector and analyzer system|

---

US5895857A|1995-11-08|1999-04-20|Csi Technology, Inc.|Machine fault detection using vibration signal peak detector|

---

SE510771C2|1996-07-05|1999-06-21|Spm Instr Ab|Procedure for evaluating the condition of a machine as well as the analyzer and apparatus for cooperation with the analyzer|

---

US5852793A|1997-02-18|1998-12-22|Dme Corporation|Method and apparatus for predictive diagnosis of moving machine parts|

---

JP3425331B2|1997-06-30|2003-07-14|株式会社東芝|Power supply|

---

US6351714B1|1998-03-03|2002-02-26|Entek Ird International Corporation|Order tracking signal sampling process|

---

FI112972B|1998-07-15|2004-02-13|Abb Research Ltd|Assessment of bearing condition|

---

US6053047A|1998-09-29|2000-04-25|Allen-Bradley Company, Llc|Determining faults in multiple bearings using one vibration sensor|

---

AUPQ152499A0|1999-07-09|1999-08-05|Commonwealth Scientific And Industrial Research Organisation|A system for monitoring acoustic emissions from a moving machine|

---

DE19938723A1|1999-08-16|2001-02-22|Busch Dieter & Co Prueftech|Signal analysis method|

---

US6351713B1|1999-12-15|2002-02-26|Swantech, L.L.C.|Distributed stress wave analysis system|

---

US6332116B1|2000-04-19|2001-12-18|National Instruments Corporation|System and method for analyzing signals of rotating machines|

---

US6591682B1|2000-08-14|2003-07-15|Pruftechnik Dieter Busch Ag|Device and process for signal analysis|

---

CA2428168A1|2000-12-01|2002-06-06|Unova Ip Corp.|Control embedded machine condition monitor|

---

US6801864B2|2001-03-13|2004-10-05|Ab Skf|System and method for analyzing vibration signals|

---

US7136794B1|2001-05-24|2006-11-14|Simmonds Precision Products, Inc.|Method and apparatus for estimating values for condition indicators|

---

TW579424B|2001-07-09|2004-03-11|Shell Int Research|Vibration analysis for predictive maintenance in machinery|

---

AT485497T|2001-12-04|2010-11-15|Skf Condition Monitoring Inc|CYCLIC TIMING FOR MACHINE MONITORING|

---

US7200519B2|2002-01-18|2007-04-03|Spm Instrument Ab|Analysis system for analyzing the condition of a machine|

---

US6618128B2|2002-01-23|2003-09-09|Csi Technology, Inc.|Optical speed sensing system|

---

US6668234B2|2002-03-22|2003-12-23|Abb Inc.|Method and apparatus for calculating the amplitude of a complex waveform associated with a rotating machine shaft after removing the running speed frequency|

---

JP3880455B2|2002-05-31|2007-02-14|中国電力株式会社|Rolling bearing remaining life diagnosis method and remaining life diagnosis apparatus|

---

US7133801B2|2002-06-07|2006-11-07|Exxon Mobil Research And Engineering Company|System and methodology for vibration analysis and condition monitoring|

---

US7243064B2|2002-11-14|2007-07-10|Verizon Business Global Llc|Signal processing of multi-channel data|

---

EP1513254A1|2003-08-26|2005-03-09|Mitsubishi Electric Information Technology Centre Europe B.V.|Filter enabling decimation of digital signals by a rational factor|

---

CN100416255C|2003-11-15|2008-09-03|西南师范大学|Torsional mechanical torque, rotation speed, angle and displacement driving sensor|

---

JP2006113002A|2004-10-18|2006-04-27|Nsk Ltd|Anomaly diagnosis system for mechanical equipment|

---

US7640139B2|2004-10-18|2009-12-29|Nsk Ltd.|Abnormality diagnosing system for mechanical equipment|

---

NZ537244A|2004-12-16|2006-10-27|Commtest Instr Ltd|Improvements in or relating to vibration analysis|

---

US7505852B2|2006-05-17|2009-03-17|Curtiss-Wright Flow Control Corporation|Probabilistic stress wave analysis system and method|

---

GB0714379D0|2007-07-21|2007-09-05|Monition Ltd|Tamping bank monitoring apparatus and method|

---

US7761256B2|2007-12-21|2010-07-20|General Electric Company|Method and system for use in analyzing vibrations of a variable speed rotating body|

---

DK2085902T3|2008-02-02|2018-03-12|Siemens Ind Software Nv|Order tracking method and system|

---

WO2010007645A1|2008-07-15|2010-01-21|グローリー株式会社|Ticket issuing device, ticket processing device, and ticket system|

---

WO2010074648A1|2008-12-22|2010-07-01|S.P.M. Instrument Ab|An analysis system|

---

DK2370801T3|2008-12-22|2018-01-22|Spm Instr Ab|Analysesystem|

---

SE535559C2|2010-01-18|2012-09-25|Spm Instr Ab|Method and apparatus for analyzing the condition of rotating part machine|WO2010128928A1|2009-05-05|2010-11-11|S.P.M. Instrument Ab|An apparatus and a method for analysing the vibration of a machine having a rotating part|

---

SE535559C2|2010-01-18|2012-09-25|Spm Instr Ab|Method and apparatus for analyzing the condition of rotating part machine|

---

GB2488092B|2010-11-03|2014-10-29|Kittiwake Developments Ltd|A sensor based means of monitoring the mechanical condition of rotating machinery that operates intermittently|

---

WO2013009258A1|2011-07-14|2013-01-17|S.P.M. Instrument Ab|A method and a system for analysing the condition of a rotating machine part|

---

EP2798325B1|2011-12-30|2019-04-03|Vestas Wind Systems A/S|Estimating and controlling loading experienced in a structure|

---

EP3239667B1|2012-01-30|2018-11-28|S.P.M. Instrument AB|An analysis system|

---

EP2626679A1|2012-02-10|2013-08-14|Siemens Aktiengesellschaft|Method for determining the damage of at least one rotatable component of a wind turbine|

---

WO2014042582A1|2012-09-11|2014-03-20|S.P.M. Instrument Ab|Apparatus for monitoring the condition of a machine|

---

WO2015074721A1|2013-11-25|2015-05-28|Aktiebolaget Skf|Bearing monitoring apparatus and method|

---

US9759213B2|2015-07-28|2017-09-12|Computational Systems, Inc.|Compressor valve health monitor|

---

US20170051682A1|2015-08-20|2017-02-23|General Electric Company|System and method for abatement of dynamic property changes with proactive diagnostics and conditioning|

---

DE102016112591A1|2016-07-08|2018-01-11|Airbus Ds Optronics Gmbh|Method for determining a degree of wear of a cooling device operated with at least one piston|

---

EP3309529B1|2016-10-11|2022-02-23|ABB Schweiz AG|Prediction of remaining useful lifetime for bearings|

---

CN106829076A|2017-02-08|2017-06-13|河南中烟工业有限责任公司|A kind of packaging facilities mechanical Fault Monitoring of HV method and apparatus chosen based on measuring point|

---

CN106995077A|2017-02-08|2017-08-01|河南中烟工业有限责任公司|A kind of packaging facilities mechanical breakdown self-diagnosing method and device|

---

DE102017107814B4|2017-04-11|2022-01-05|Phoenix Contact Gmbh & Co. Kg|Condition monitoring device for monitoring the condition of a mechanical machine component|

---

DE102017109460A1|2017-05-03|2018-11-08|Prüftechnik Dieter Busch AG|System and method for vibration measurement on a machine|

---

DE102017110342A1|2017-05-12|2018-11-15|Prüftechnik Dieter Busch AG|RMS value determination of a machine vibration quantity|

---

US10655607B2|2017-06-02|2020-05-19|General Electric Company|Systems and methods for detecting damage in wind turbine bearings|

---

CN109923486B|2017-08-11|2022-02-18|李荣圭|Accurate predictive maintenance method for driving part|

---

KR102039742B1|2017-08-11|2019-11-01|아이티공간|Predictive maintenance method of driving part|

---

KR102103143B1|2018-03-14|2020-04-22|아이티공간|Predictive maintenance method of driving device|

---

CN108844742B|2018-09-06|2020-08-18|国电联合动力技术有限公司|Method and system for monitoring lubricating state of generator bearing of wind turbine generator|

---

CN110067767B|2019-04-25|2021-04-13|沈阳鼓风机集团自动控制系统工程有限公司|Method and device for monitoring state of centrifugal compressor unit|

---

DE102019219772A1|2019-09-26|2021-04-01|Robert Bosch Gmbh|Sensor system, linear device and method for a sensor system|

---

US10931115B1|2019-09-30|2021-02-23|General Electric Company|Electrical power systems having a cluster transformer with multiple primary windings|

---

CN110987427B|2019-12-31|2021-11-09|安徽容知日新科技股份有限公司|Data processing method, device and system for mechanical equipment|

法律状态:

优先权:

---

申请号 | 申请日 | 专利标题

---

SE1000045|2010-01-18|

---

SE1000631A|SE535559C2|2010-01-18|2010-06-11|Method and apparatus for analyzing the condition of rotating part machine|SE1000631A| SE535559C2|2010-01-18|2010-06-11|Method and apparatus for analyzing the condition of rotating part machine|

---

CN201180006321.7A| CN102822644B|2010-01-18|2011-01-13|For analyzing the equipment of the state of the machine with rotary part|

---

EP11733155.3A| EP2526389A4|2010-01-18|2011-01-13|Apparatus for analysing the condition of a machine having a rotating part|

---

US13/522,023| US9279715B2|2010-01-18|2011-01-13|Apparatus for analysing the condition of a machine having a rotating part|

---

EA201290660A| EA024339B1|2010-01-18|2011-01-13|Apparatus for analysing the condition of a machine having a rotating part|

---

PCT/SE2011/050035| WO2011087440A1|2010-01-18|2011-01-13|Apparatus for analysing the condition of a machine having a rotating part|

---

US15/010,109| US10330523B2|2010-01-18|2016-01-29|Apparatus for analysing the condition of a machine having a rotating part|

---

US16/435,208| US20190391004A1|2010-01-18|2019-06-07|Apparatus for analysing the condition of a machine having a rotating part|