• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    The present invention relates to a method for monitoring an aircraft engine (1) (2) operating in a given environment, characterized in that it comprises the implementation by data processing means (31) of steps of: (a) Receiving a sequence of n-tuples (x1_exec ... xn_exec; yexec) of physical parameter values relating to said aircraft engine (1) (2), including at least one inherent endogenous parameter the operation of the engine (1) and at least one exogenous parameter specific to said environment; (b) For each tuple (x1_exec ... xn_exec; yexec) of the received sequence, computation based on a regression model of a normalized value (yexec_norm) of the endogenous parameter with respect to the exogenous parameters; (c) identifying at least one stabilized phase in said normalized n-tuplet sequence (x1_exec ... Xn_exec; yexec_norm) from a set of phase signatures; (d) For each stabilized phase, calculation of mean values of the physical parameters on the part of the sequence of n-tuples (x1_exec ... xn_exec; yexec) corresponding to the stabilized phase, so as to obtain a tuple defining a recurrent point of said flight of the aircraft (2).
    公开号:FR3028331A1
    申请号:FR1460853
    申请日:2014-11-10
    公开日:2016-05-13
    发明作者:Aurelie Gouby；Guillaume Bothier
    申请人:SNECMA SAS；
    IPC主号:
    专利说明:

    [0001] GENERAL TECHNICAL FIELD The present invention relates to the field of "Health Monitoring" of aeronautical equipment.
    [0002] More specifically, it relates to a method of monitoring an aircraft engine operating in a given environment. STATE OF THE ART The Health Monitoring (French health status monitoring) designates the monitoring of the evolution of health and the state of equipment, in particular a turbomachine, throughout its life . One of the objectives of the Health Monitoring is to anticipate and plan the maintenance operations in a sufficiently relevant way to avoid any problem that could cause a malfunction or even a breakdown (with probably dramatic consequences if it takes place in the air). To this end, we seek to finely track all accessible engine parameters to predict preventive maintenance operations rather than curative (which are particularly expensive).
    [0003] The implementation of the engine monitoring requires a detailed expertise of the operation of the engine according to its own operating parameters (so-called "endogenous" parameters, for example the inlet pressure of the combustion chamber, the temperature of the exhaust gas , etc.) but also external parameters, related to the environment (so-called "exogenous" parameters, for example the altitude or the temperature of the incoming air). It has been proposed to analyze the behavior of an engine in comparison with the behavior encountered in the past (see patent application FR2971595). To do this, it is necessary to be able to compare the engines with each other, that is to say that the influence of the context on the studied parameters must be eliminated, in other words to normalize them (for example, sub-Saharan an exhaust gas temperature which will be higher than normal, without the engine functioning being abnormal). These normalization methods, which work by learning a model, are built with a precise point and representative of the flight: the "snapshot" (in French "instantaneous"). These representative points of the flight will be called thereafter recurring points flight by flight. If for civil engines, where the snapshot phase is typically a cruise phase, these methods are entirely satisfactory, the situation is more complicated for military engines. Indeed, the flight profile is different: a reconnaissance aircraft constantly changes its speed and altitude. Modeling via a "medium" operation thus makes it very difficult to take into account the operating conditions of an engine of a fighter jet that can be pushed to its limits in these very severe external conditions during certain missions. therefore desirable to have a reliable, efficient, and reproducible way of monitoring the operation of an aircraft engine including military, regardless of the variety, diversity and extent of its operations. PRESENTATION OF THE INVENTION The present invention proposes according to a first aspect a method of monitoring an aircraft engine operating in a given environment, characterized in that it comprises the implementation by data processing means of steps of: (a) Receiving a sequence of n-tuples of physical parameter values relating to said aircraft engine, including at least one endogenous parameter specific to the operation of the engine and at least one exogenous parameter specific to said environment , said values being measured over time by sensors so that each tuple of the sequence defines a point of a flight of said aircraft, a set of reference sequences of n-tuples of values of said physical parameters being stored in a database stored on data storage means; (b) For each tuple of the received sequence, computed as a function of a regression model associated with a subset of said set of tuple reference sequences, of a normalized value of the endogenous parameter with respect to exogenous parameters, so as to obtain a sequence of standardized n-tuples; (c) identifying at least one stabilized phase in said normalized n-tuplet sequence from a set of phase signatures stored in said data storage means database, each signature being defined by a uplet of values of the physical parameters and a tuple of associated variance values, a stabilized phase corresponding to a portion of said sequence representative of a flight time greater than a given threshold and in which the values of the standardized n-tuples coincide with the values of the tuple of a signature with the variance associated with it; (d) For each stabilized phase, calculation of average values of the physical parameters on the part of the sequence of n-tuples corresponding to the stabilized phase, so as to obtain a tuple defining a recurrent point of said flight of the aircraft, and transmission to interface means.
    [0004] The normalization of endogenous parameters makes it possible to dispense with the context and to be able to compare two flights, and the use of signatures makes it possible to find stabilized phases and hence recurring flight-to-flight points characterizing easily and efficiently the behavior of the engine compared to to known behaviors. According to other advantageous and nonlimiting characteristics: step (a) comprises the separation of the values of the exogenous parameters and the values of the endogenous parameters; step (b) comprises, for each tuple of the received sequence, a preliminary step (b0) of determining an exogenous class of the tuple according to the values of the exogenous parameters of said tuple, among a a plurality of exogenous classes each defined by the exogenous parameter values of a subset of said set of n-tuplet reference sequences; step (b0) comprises projecting said tuple in the context classes so as to identify the closest exogenous class according to a distance criterion; the regression model used in step (b) for a tuple of the received sequence is the model associated with the exogenous class determined for said tuple; the method comprises a preliminary phase of processing said set of reference sequences of n-tuples of the database, comprising the implementation by means of processing data of steps of: (a0) classification of the reference sequences n-tuples so as to generate said plurality of exogenous classes; (ai) for each exogenous class, determining said regression model associated with the subset of said set of n-tuplet reference sequences by a regression modeling the value of the endogenous parameter as a function of the exogenous parameter values from the set n-tuples of the exogenous class; (a2) for each tuple of the set of reference sequences, calculating for the exogenous class of the n-tuple an estimated value of the endogenous parameter and an associated residue; (a3) For each tuple of the set of reference sequences, calculating the normalized value of the endogenous parameter with respect to the exogenous parameters, so as to obtain a set of standardized n-tuplet reference sequences; (a4) for each standard n-tuplet reference sequence, identifying at least one stabilized phase in said normalized n-tuplet reference sequence, a stabilized phase corresponding to a portion of said representative time-of-flight sequence greater than a given threshold and in which normalized n-tuplet values are constant at a given variance; (a5) for each stabilized phase of each reference sequence, calculating average values of the physical parameters on the part of the sequence of n-tuples corresponding to the stabilized phase, so as to obtain a tuple defining a potential recurrent point of a flight of the aircraft; (a6) classifying n-tuples defining a potential recurrent point obtained to generate a plurality of cells, each associated with a subset of the n-tuples defining a potential recurring point; (a7) for at least one generated cell, calculating a n-tuple of values of the physical parameters and a tuple of associated variance values so as to define a phase signature associated with the cell, and storing the signatures of phases in said database means for storing data. the classification of steps (a0) and (a6) is carried out according to a non-supervised method chosen from the k-means algorithm, the Kohonen self-adaptive map method and the hierarchical ascending classification (CAH) ; step (a6) comprises, for each cell generated, calculating a density of the cell defined as the number of reference sequences for which the cell comprises at least one n-tuple defining a potential recurrent point obtained, step (a7) being carried out for the cells having the highest density; each endogenous parameter is chosen from a pressure at the output of an engine booster, a static pressure at the inlet of a combustion chamber of the engine, a temperature at the output of the engine booster, a temperature of the exhaust gases of the engine, motor, a mass flow of fuel at the inlet of a high-pressure compressor of the engine, and a high-pressure rotation speed at the inlet of the compressor High-pressure of the engine; each exogenous parameter is chosen from an altitude, a temperature at the inlet of an engine blower, and a low-pressure rotational speed at the inlet of the engine blower; the sensors are integrated into the engine, said sequence of n-tuples being received from the engine via an Aircraft Communication Addressing and Reporting System (ACARS) transmission; the n-tuples of a sequence are measured by the sensors at a regular frequency of between 0.1 Hz and 10 Hz; the method comprises adding the received sequence of n-tuples to said set of n-tuplet reference sequences. According to a second aspect, the invention relates to equipment for monitoring an aircraft engine operating in a given environment, comprising: data processing means; data storage means storing in a database; data: o a set of reference sequences of n-tuples of physical parameter values relating to said aircraft engine, including at least one endogenous parameter specific to the operation of the engine and at least one exogenous parameter specific to said environment, and o a a set of phase signatures, each signature being defined by a n-tuple of values of the physical parameters and an n-tuple of associated variance values, - interface means, the equipment being characterized in that the means of data processing are configured to implement: a module for receiving a sequence of n-tuples of values of the physical parameters relating to said aircraft engine, these values being measured over time by sensors so that each tuple of the sequence defines a point of a flight of said aircraft; A computation module, for each n-tuple of the received sequence, according to a regression model associated with a subset of said set of n-tuple reference sequences, of a normalized value of the endogenous parameter by relative to the exogenous parameters, so as to obtain a sequence of standardized n-tuples; A module for identifying at least one stabilized phase in said sequence of n-tuples normalized from said set of phase signatures, a stabilized phase corresponding to a part of said sequence representative of a flight time greater than one given threshold and in which normalized tuple values coincide with tuple values of a signature with the associated variance; A calculation module, for each stabilized phase, of average values of the physical parameters on the part of the sequence of n-tuples corresponding to the stabilized phase, so as to obtain a tuple defining a recurrent point of said flight of the aircraft, and a transmission module for the interface means. According to other advantageous and nonlimiting features: the data processing module is furthermore configured to implement: a module for classifying reference sequences of n-tuples 30 so as to generate a plurality of exogenous classes; each defined by the exogenous parameter values of a subset of said set of n-tuplet reference sequences; A module for determining, for each exogenous class, said regression model associated with the subset of said set of n-tuplet reference sequences by a regression modeling the value y of the endogenous parameter as a function of the values of the exogenous parameters from the set of n-tuples of the exogenous class; A calculation module, for each tuple of the set of reference sequences, of an estimated value of the endogenous parameter and of an associated residual for the exogenous class of the tuple; A calculation module, for each tuple of the set of reference sequences, of the normalized value of the endogenous parameter with respect to the exogenous parameters, so as to obtain a set of standardized n-tuple reference sequences; An identification module, for each standard n-tuplet reference sequence, of at least one stabilized phase in said normalized n-tuplet reference sequence, a stabilized phase corresponding to a part of said representative sequence of a flight time above a given threshold and in which normalized tuple values are constant at a given variance; A calculation module, for each stabilized phase of each reference sequence, of average values of the physical parameters on the part of the sequence of n-tuples corresponding to the stabilized phase, so as to obtain a tuple defining a recurrent point; potential of a flight of the aircraft; A module for classifying n-tuples defining a potential recurrent point obtained so as to generate a plurality of cells, each associated with a subset of the n-tuples defining a potential recurring point; A calculation module, for at least one generated cell, of a n-tuple of values of the physical parameters and an n-tuple of associated variance values so as to define a phase signature associated with the cell, and storage phase signatures in said database of data storage means. According to a third aspect, the invention relates to a system comprising: an aircraft comprising a motor and sensors; - Equipment according to the second aspect of the invention for monitoring an aircraft engine operating in a given environment.
    [0005] According to a fourth and a fifth aspect, the invention relates to a computer program product comprising code instructions for the execution of a method according to the first aspect of the invention for monitoring an operating aircraft engine. in a given environment; and computer-readable storage means on which a computer program product comprises code instructions for executing a method according to the first aspect of the invention for monitoring an aircraft engine in operation in a given environment.
    [0006] PRESENTATION OF THE FIGURES Other features and advantages of the present invention will appear on reading the description which follows of a preferred embodiment. This description will be given with reference to the accompanying drawings, in which: FIG. 1 represents an example of an environment in which the method according to the invention is implemented; FIGS. 2a-2b illustrate the steps of two phases of an example of the method according to the invention; FIG. 3 represents an example of exogenous classes used in a process according to the invention; FIG. 4 represents examples of stabilized phases identified during the implementation of the method according to the invention; FIGS. 5a-5b show examples of cells used during the implementation of the method according to the invention.
    [0007] DETAILED DESCRIPTION With reference to FIG. 1, the present method is a method of monitoring an aircraft engine 1 operating in a given environment, in particular a military aircraft (for example a fighter jet) on mission. The engine 1 is typically all or part of a turbomachine, in particular double flow. The objective is to find flight phases where the behavior of the engine is identical regardless of the context, and to determine snapshots, in other words the "recurring flight-to-flight points" mentioned above, associated with these phases. These recurring points make it possible to refer to behaviors already encountered in the past and thus to make it easier to plan maintenance operations. The present method can be applied to any measurement monitoring, but preferably it is a "pseudo real-time" monitoring: the engine 1 is equipped with sensors 20, active during the flight of the engine. 2. This aircraft then sends regularly to the ground small instant messages including the values of the measurements from the sensors 20. These messages are sent for example by satellite 35 (ACARS protocol) through transmission means, and equipment 3 disposed on the ground comprising data processing means 31 (for example a processor) and data storage means 32 (for example a hard disk) receives the data contained in these messages via a base station 34 and processes them for setting implementation of the method.
    [0008] Those skilled in the art will understand that the latter is not limited to any procedure for transmitting measurements to the equipment 3 (for example, it is possible for the measurements to be stored on the aircraft during flight time, and transmitted en bloc. to equipment 3 after landing). Moreover, the treatment can be deferred in time. It is even conceivable that the equipment 3 is integrated with the aircraft 2. The equipment 3 (or other equipment) is equipped with interface means 33 (such as a keyboard and a screen) for interacting, and particular for the display of results (see below). In general, the first step (a) of the present method consists in the reception by the data processing means 31 of a sequence of n-tuples of values xi_ ex 'X: exec, V exec physical relating to said engine 1 d aircraft 2, including at least one endogenous parameter specific to the operation of the engine 1 and at least one exogenous parameter specific to said environment. By "tuple" is meant a vector comprising a value for each of the parameters. The values are measured over time by the sensors 20, and each tuple is associated with a time instant. A sequence designates a flight of the aircraft, and the n-tuples of the sequence are points of the flight, thus obtained consecutively during the flight. Preferably, the values are acquired (and possibly emitted) at regular time intervals, for example at a frequency between 0.1 Hz and 10 Hz, in particular about 1 Hz (a value acquired for each parameter every second of the flight). Endogenous or exogenous parameters are physical parameters. They thus represent physical quantities such as a temperature or a pressure. Those skilled in the art will choose the type of physical quantity to be measured according to the effects to be monitored on the engine. For each parameter, the associated sensor 20 is adapted to the size (thermometer, pressure gauge, etc.). As explained, some physical parameters are "endogenous", and are therefore specific to the operation of the engine 1. In other words, they are parameters whose value is directly impacted by the operation of the engine. For example, in the case of a dual-flow military motor (with each time the associated code in parentheses): - the pressure at the output of the booster (P23); - the static pressure at the inlet of the combustion chamber (PS32); the temperature at the output of the booster (T23); - the temperature of the exhaust gases (TM49); the mass flow of fuel at the inlet of the high-pressure compressor (W25); - the high-pressure rotation speed at the inlet of the high-pressure compressor (XN25). The other physical parameters are "endogenous", and are therefore specific to the environment of the engine 1, i.e. the context. In other words, they are parameters whose value is not affected by the operation of the engine, but that the engine experiences. For example, in the case of a dual-flow military engine (each with the associated code in parentheses): the altitude (ALTF); - The temperature at the inlet of the blower (T2); - The low pressure rotation speed at the inlet of the fan (XN2). An exogenous parameter (x value) is an "explanatory" or "predictive" variable, as opposed to an endogenous parameter (y value) that is an "explain" or "predict" variable. In other words, the x value of the exogenous parameter is a cause, when the y value of the endogenous parameter is a consequence. A tuple x1 ... xn; y designates a point acquisition: for x values of the exogenous parameters, y values of the endogenous parameters are measured. In the remainder of the present description, we will take the example of an endogenous parameter (and n exogenous parameters), but it will be understood that it is sufficient to repeat the process steps for each endogenous parameter if there are several, and that the number of exogenous parameters does not matter. In the examples, three exogenous parameters will be taken. A set of reference sequences of n-tuples --- xni; i) iE [[1, p]] values of said physical parameters (p sequences) 5 is stored in a database itself stored on the data storage means 32. Each sequence corresponds to a flight of an aircraft similar (with a similar engine), and for each of its flights we have a sequence of n-tuples. The n-tuples of the base each define reference values yi of the endogenous parameters for xi values of the exogenous parameters. Sequences and "reference" values are understood to mean that they are acquired during known prior flights, that is, they can be considered exploitable. Any abnormal values have already been removed from the database. The received sequence of Xiexec xnexec n-tuplets; Yexec designates the "monitored" sequence, that is to say that of the flight for which it is sought to monitor the engine 1 operating in a given environment. As explained, this monitored sequence may be a sequence obtained in real time (in particularly in a ACARS operation) or a sequence obtained in non-real time (executables X1 exec-Xn exec; Yexec stored in the database and placed on hold). Learning Phase The present process comprises two phases. The first is a learning phase and the second is an execution phase. Preferably, the learning phase is implemented beforehand so as to create the models that will be described later (and if necessary store them on the data storage means 32), and the execution phase is then implemented at each new reception of a sequence. This is the execution phase that allows the monitoring of the engine 1, object of the invention. The learning phase can be restarted from time to time to update the models. Alternatively, it is quite possible not to make prior learning and determine the models piecemeal at each implementation of the execution phase. The two phases will be described in the following description. The learning phase can be viewed as a set of steps of processing only the data of the database (ie, regardless of the sequence of n-tuples x1 ex 'exec; exec) - With reference to FIG. of learning begins with a step (a0) of classification of the reference sequences of n-tuples --- xni; i) iE [[1, p]] so as to generate said plurality of exogenous classes. Since the flight contexts of a military aircraft are very varied, it is indeed important to be able to classify them. Each "exogenous class" is defined by the values (xli Xni) iElitm (pj designates the cardinal of the exogenous class), the sum of the pj values p) exogenous parameters of a subset of said set of reference sequences of n -uplets --- xni; i) iE [[1, p]] - In other words, this step consists of the partition 20 of the tuples ... xni; yi) ielitpil deprived of values of endogenous parameters (ie their restriction to all or part of the exogenous parameters, in this case hyperplanes in our example to a single endogenous parameter) via unsupervised classification methods (in particular chosen from the k-means algorithm, the Kohonen self-adaptive map method and the hierarchical ascending classification (CAH)). To rephrase again, we place each of the flight points defined by a tuple xni; yi) ielitpil in a restricted subspace (in terms of dimension) to the exogenous parameters, and we partition this space into k classes (for example between 2 and 10), each of which will be representative of a general type of context in which the exogenous parameters will have similar values between two flights (for example "low altitude in Qatar"). Each exogenous class is thus associated with a subset of the set of reference sequences of n-tuples (x1, ... x72,; y,), Eli1, m, subset consisting of "close" flights in terms of context.
    [0009] The advantage of setting k for example between 2 and 10 is to define a maximum number of exogenous classes for the classification to be relevant (do not have as many classes as points when they are all far away.). An optimization method makes it possible to choose k so as to maximize the difference between the different groups and to minimize the difference between points of the same class). FIG. 3 thus illustrates an example representing the arrangement of the tuples (x1, ... xm; y,), Elitpil in a three-dimensional space generated by the three exogenous parameters ALTF, T2 and XN2. We clearly see three exogenous classes that stand out.
    [0010] This step is followed by a step (a1) of determining a plurality of regression models each associated with a subset of said set of reference sequences of n-tuples (x1, ... xm; (in particular a class exogenous), by a regression modeling y as a function of x on the values of the n-tuples (x1, ... xm; yi), Elitpil of the subset (in other words, those associated with the exogenous class). In the case where there are several endogenous parameters, this step is repeated so as to determine regression models modeling each y as a function of x.These regression models will be used in the execution phase. statistical methods well known to those skilled in the art for analyzing the relationship of a variable (here y) with respect to one or more others (here x1 .. xn), step (a1) consists of other terms to determine functions fi (j is an index denoting the j-th exogenous tired) of allowing to best approximate the values y, according to the values xl, x7, for a given type of link. Thus, linear, polynomial, exponential, logarithmic regressions are known. The choice of the type of link used is advantageously made according to the shape of the curve and can be done automatically by optimization by maximizing a coefficient of determination, for example as described in the patent application FR2939928. In a third step (a2), the learning phase comprises the calculation (repeated for each endogenous parameter), for each of the tuple ... xni; yi) ielitm, of an estimated value of the endogenous parameter 10 and an associated residue resii, for the class j to which the n-tuple --- xni belongs; Yi) iE [[1, p]] - The residual is the difference between an estimated value and a measured value. From the regression model, we obtain these values simply by the formulas = xni), and resii = yi - 9f ', for E pj]]. Once the regression model has been created, the data processing means 31 deduce another model used in the execution phase: this is the normalization model. As explained, the objective is to remove the influence of contexts on the endogenous parameters, in other words to make them normalized and comparable since they are brought back to the same flight conditions. In step (a3), is thus calculated for each n-tuple --- xni; Yi) iE [[1, p]] of the set of reference sequences, a normalized yiinorm value of the endogenous parameter with respect to the exogenous parameters, so as to obtain a set of normalized n-uplet reference sequences (x ... xni; v norm) iE [[1, p]] - The normalization model associates with each value of an endogenous parameter the normalized value of this parameter. The normalization model is for example given by the formula v norm = 371 + resii, where 371 is the average value of the parameter y for the tuples of the exogenous class j. Indeed, since is the prediction made from the exogenous variables and that resii = yi -, the application of the preceding formula makes it possible to subtract the influence of the endogenous, and there remains only one signal centered in the average and whose variability can no longer be explained by the context.
    [0011] Once the endogenous parameters are normalized, the stabilized phases will be searched during the flight. Thus, in a step (a4), at least one stabilized phase in said reference sequence is identified (if possible) for each reference sequence of normalized n-tuples (xli xni; v norm) iE [[1, p]] normalized n-tuplets (--- xni; Yi norm) iE [[1, p]] a stabilized phase corresponding to a part of said sequence representative of a flight time greater than a given threshold, for example ten minutes, (if the acquisition frequency is constant, then this time threshold corresponds to a threshold of number of consecutive points) and in which the values of the normalized n-tuples (x1 Yi norm) iE [[1, p]] are constant to a given variance. In other words, a minimum stabilized phase duration t to be attained and for each parameter a tolerance on the variance is fixed. It should be noted that a military aircraft pilot often remains in stable phases for a short period of time, which is why it is desirable to test several minimum time thresholds and several variance window sizes. In a known manner, the variance is calculated as the average of the squared residuals. Figure 4 illustrates the value sequences for three endogenous (normalized) parameters P23, TM49 and XN25. Three stabilized phases are identified in which each of the endogenous parameters has a substantially constant value. Once the stabilized phases found in the different flights of learning (ie for each sequence of the set), we summarize each phase (step (a5)) in a tuple by averaging the parameter values on the phase (ie on the n-tuples 30 (--- xni; Yi norm) iE [[1, p]] of said part of sequence corresponding to the phase). For a flight, we obtain a matrix of size nxk where n is the number of parameters and k the number of stabilized phases found during the flight. The tuple 37 defines a "potential" recurring point of a flight of the aircraft 2, and thus potentially constitutes a "signature" of the phase.
    [0012] In the next step (a6), the tuples .1,2; y defining a potential recurrent point obtained are classified by the data processing means 31 in the manner of what is done in step (a0). As before, a plurality of unsupervised methods can be implemented, and preferably the self-adaptive map method of Kohonen is chosen. The classification makes it possible to generate a plurality of cells according to a map as represented in FIG. 5a, each associated with a subset of the n-tuples ..... 37 defining a potential recurring point.
    [0013] This step advantageously comprises, for each cell generated, the calculation of a "density" of the cell defined as the number of reference sequences for which the cell comprises at least one n-tuple defining a potential recurrent point obtained therefrom. In other words, we calculate the number of potential recurring points per cell, counting only one point if a flight (ie a sequence) comprises several times the same phase). The number displayed on each cell of FIG. 5a is thus the number of flights projected in each cell). The densest cell or cells are identified (the cell at 184 flights in the example of Figure 5a), and selected. Alternatively, a minimum threshold of cell density can be provided. In any case, if the base of n-tuples is not large enough to construct a relevant map, it can be expected to group some neighboring classes into "meta-classes", as shown in Figure 5b, where build a map with fewer cells. For at least one generated cell (especially those with the highest density, i.e. those where the dots are truly "recurrent"), a signature of the cell is calculated in step (a7). This signature is a signature of a type of recurrent phase of the flights of the aircraft. A signature s is defined by a tuple xls x '; ys of values of the physical parameters and a tuple vars (xi) vars (x 72); vars (y) of associated variance values. The tuple xls ... x '; ys of values is typically the "representative" of the cell, that is, the center of the cell in the sense of Kohonen, where the mean of the n-tuples y of the cell. The variance is the "real" variance of the cell parameters, that is typically that vars (x) = (x fc - x) 2, where m is the number of n-tuples in the cell ( we note that this formula is the same for endogenous variables y). The obtained phase signatures are stored in said database of data storage means (32).
    [0014] Execution phase As explained above, the learning phase represents a preparatory work to accelerate the execution phase (which corresponds to the core of the present method according to the invention). The learning phase can alternatively be performed "at the same time" as the execution phase. In the description of this part, reference will be made to all the associated formulas described above. This phase makes it possible to monitor the engine 1 operating in a given environment during a flight defined by a sequence of n-tuples xi exec --- xn exec; Y exec of values of the physical parameters of the engine 1. This phase is illustrated by FIG. 2b. If the learning phase has been implemented previously, the models and signatures can be loaded from the data storage means 32, as illustrated in this figure.
    [0015] Step (a), already mentioned, sees the receipt of the X1 exec executes exec executable. This step includes the separation of the Ex executable parameters XX exec and Ex endogenous parameters exx parameters (for example by means of a list).
    [0016] In a step (b), for each point of the sequence, a value of each endogenous parameter is calculated for the values .X1 exec ... Xn exec of the exogenous parameters as a function of a regression model associated with a subset of said parameter. set of reference sequences of n-tuples --- xni; i) iE [[1, p]] (in this case the model associated with an exogenous class, determined if necessary in the learning phase). For this, step (b) comprises a preliminary step (b0) of determining an exogenous class of the tuple as a function of the values .X1 exec ... Xn exec exogenous parameters of said tuple, among a plurality exogenous classes each defined by the (./Cli Xni) iElitm values of the exogenous parameters of a subset of said set of reference sequences of n-tuples --- xni; i) iE [[1, p]] - In this case, the exogenous class of the tuple is typically the one of which it is the closest (according to its exogenous parameters). For this, this determination is done for example by projecting said tuple in the exogenous classes so as to identify the closest exogenous class according to a distance criterion (conventional method in which the data processing means 31 calculate for each exogenous class a distance (according to a given distance criterion) between the class and a restriction .X1 exec ... Xn exec to the exogenous parameters of said tuple, and 25 the class j for which the distance is the shortest is chosen) Once the exogenous class has been determined, the calculation is done as previously explained by first calculating an estimated value estimated with the endogenous parameter xec = fi (Xi exec ... Xn exec). The associated Resexec residue is then computed: 30 reseixec = Yexec 9eixec- Finally, the normalized value is obtained directly from the formula v exec norm = + resejxec.
    [0017] We then obtain a sequence of standardized n-tuples X1 exec --- Xn exec; Yexec norm, i.e. for which the influence of the context has been suppressed. In step (c), the stabilized phases are identified from the set of phase signatures stored in said data storage means database 32. As explained, a stabilized phase corresponds to a portion of said sequence representative of a time of flight greater than a given threshold and in which the values of the normalized n-tuples xl exec --- xn exec; Yexec norm coincide with the values of the tuple --- xns; ys a signature with the associated variance. Concretely, if for a duration of at least the given threshold, the difference between each parameter xi_ exec Xn exec,: V exec norm of the standardized tuple and each corresponding parameter of the tuple xls ... xns; ys of a signature s, at the corresponding value var (x1) vars (xn); vars (y) of the n-tuple of variance associated with the signature s, then it is indeed a known stabilized phase, and a recurring flight-to-flight point can be identified. In step (d), these recurring points are defined as a tuple of average values of the physical parameters on the part of the sequence of n-tuples Xiexec ... xn exec; Yexec norm stabilized phase. These recurring points are then transmitted to interface means 33 for operation (and / or for example stored on the storage means 32), for example determining maintenance steps to be provided, using known models. It should be noted that the received sequence of execXn n-tuples x1: Yexec exec, can then be added to said set of reference sequences of tuples (x1, ... xni; ypielitpil (to reinforce the learning base and to refine the models).
    [0018] Equipment and system corresponding to the equipment 3 (shown in FIG. 1) for carrying out the process just described (monitoring an aircraft engine 1 operating in a given environment) comprises data processing means 31, data storage means 32, and interface means 33. The data storage means 32 store in a database: o The set of reference sequences of n-tuples - --xni; Y i) iE [[1, p]] of physical parameter values relating to said aircraft engine 1, and o The set of phase signatures, each signature being defined by a tuple xls xns; ys of values of the physical parameters and a tuple vars (xi) vars (xn); vars (y) of associated variance values, the data processing means 31 are configured to implement: a module for receiving a sequence of n-tuples xl exec --- xn exec; Yexec of parameter values physical relative to said aircraft engine 1, said values being measured over time by sensors 20 so that each tuple of the sequence defines a point of a flight of said aircraft (2); - A calculation module, for each tuple x1 exec ... Xn exec; Yexec of the received sequence, according to a regression model associated with a subset of said set of n-tuplet reference sequences (x1, ... xm;, of a normalized value Yexec norm of the endogenous parameter by relative to the exogenous parameters, so as to obtain a sequence of normalized n-tuples xl exec --- xn exec; Yexec norm - - An identification module of at least one stabilized phase in said normalized n-tuplet sequence Xi exec ... Xn exec; Yexec norm from said set of phase signatures, a stabilized phase corresponding to a portion of said sequence representative of a flight time greater than a given threshold and in which the values of the normalized n-tuples xi_ ex 'Xn exec; Yexec norm coincide with the values of the tuple xls xns; ys of a signature with the associated variance; - A calculation module, for each stabilized phase, of mean values of the physical parameters on the part of the sequence of n -uplets Xiexec ... Xn exec, Yexec norm corresponding to the stabilized phase, so as to obtain a .1,2 tuple; defining there a recurrent point of said flight of the aircraft 2; and a module for transmitting recurring points to the interface means 33.
    [0019] In a preferred manner, the data processing means 31 also implement a determination module, for each exec nt-Xn exec; Yexec of the received sequence, of an exogenous class of the tuplet as a function of the values Xi. exec Xn exec exogenous parameters of said tuple, among a plurality of exogenous classes each defined by the values (./C1, Xm) tElitpfl exogenous parameters of a subset of said set of reference sequences of n -uplets (x1, ... Xm; yi) / Elitm. The regression model associated with a subset of said set of reference sequences of n-tuples (x1, ... xm; yi), Elitpil is then the model associated with the exogenous class.
    [0020] This determination module can realize the projection of said tuple in the context classes so as to identify the closest exogenous class according to a distance criterion. If the equipment 3 also implements the learning phase, then the data processing module 31 is further configured to implement: a module for classifying the reference sequences of n-tuples --- xi ; Yi) iE [[1, p1 so as to generate a plurality of exogenous classes each defined by the values (xli Xni) ielitm of the exogenous parameters of a subset of said set of reference sequences of n-tuples Xni; ; A module for determining, for each exogenous class, said regression model associated with the subset of said set of reference sequences of n-tuples xni; yi) ielitpil by a regression modeling the value y of the endogenous parameter as a function of the values x1 ... xn of the exogenous parameters from the set of n-tuples of the exogenous class; - a calculation module, for each tuple xni; yi) ielitpil of the set of reference sequences, an estimated value of the endogenous parameter and an associated residue resii for the exogenous class j of the n-tuple (x1, ... Xni;; - a calculation module for each tuple (x1, ... xni; yi), Elitpil of the set of reference sequences, of the norm normalized value of the endogenous parameter with respect to the exogenous parameters, so as to obtain a set of sequences of reference of standardized n-tuples (xli xni; y n01.77) jE1113) 11; An identification module, for each reference sequence of normalized tuples (xli ... xni; v norm) iE [[1, p]], of at least one stabilized phase in said reference sequence of n-tuples standardized ... xni; Yi norm) iE [[1, p]] a stabilized phase corresponding to a part of said sequence representative of a flight time greater than a given threshold and in which the values of the normalized n-tuples (x1i --- Xni; Yi norm) .E [[1, p]] are constant at a given variance; A calculation module, for each stabilized phase of each reference sequence, of average values of the physical parameters on the part of the sequence of n-tuples --- xni; Yi norm) iE [[1, p]] corresponding to the stabilized phase, so as to obtain a tuple; defining a potential recurrent point of a flight of the aircraft 2; A module for classifying n-tuples defining a potential recurrent point obtained so as to generate a plurality of cells, each associated with a subset of the n-tuples defining a potential recurring point; A calculation module, for at least one generated cell, of a tuple xis --- Xns; Ys of values of physical parameters and a tuple vars (x1) vars (xn); vars (y) associated variance values so as to define a phase signature associated with the cell, and storing phase signatures in said database of the data storage means 32.
    [0021] The equipment 3 fits as explained in a preferred manner in a system further comprising the aircraft 2 (preferably military) comprising the engine 1 and the sensors 20 measuring the values X1 exec --- Xn exec; Yexec the physical parameters of the motor 1.
    [0022] Computer program product According to a fourth and a fifth aspect, the invention relates to a computer program product comprising code instructions for execution (on data processing means 31, in particular those of the equipment 3 ) a method according to the first aspect of the invention for monitoring an aircraft engine 1 operating in a given environment, as well as storage means readable by a computer equipment (for example the storage means data 32 of this equipment 3) on which we find this product computer program.
    权利要求:
    Claims (13)
    [0001]
    REVENDICATIONS1. A method of monitoring an aircraft engine (1) operating in a given environment, characterized in that it comprises the use by data processing means (31) of steps of: a) Receiving a sequence of n-tuples ex exec (Yexec) of physical parameter values relating to said aircraft engine (1) (2), including at least one endogenous parameter specific to the operation of the engine (1) and at least one exogenous parameter specific to said environment, said values being measured over time by sensors (20) so that each tuple of the sequence defines a point of a flight of said aircraft (2), a set of n-tuplet reference sequences (x1, ... xm; yi), Elitpil of values of said physical parameters being stored in a database stored on data storage means (32); (b) For each exec tuple ... Xn exec; Yexec) of the received sequence, calculated according to a regression model associated with a subset of said set of reference sequences of n-tuples --- xni; Yi) tEL [tp]], of a normalized value (v, exec norm) of the endogenous parameter with respect to the exogenous parameters, so as to obtain a sequence of standardized n-tuples (x1 exec ... Xn exec; Yexec norm ); (c) identifying at least one stabilized phase in said normalized n-tuplet sequence (Xi exec ... Xn exec; Yexec norm) from a set of phase signatures stored in said database of means for storing data (32), each signature being defined by a tuple (x1, ... xns; ys) of values of the physical parameters and a tuple (vars (xi) vars (xn); vars (y) ) of associated variance values, a stabilized phase corresponding to a portion of said sequence representative of a time of flight above a given threshold and in which the values of the normalized n-tuples (xi_ ex 'exec; Yexec norm) coincide with the tuple values (x1, ... xns; ys) of a signature with the associated variance; (d) For each stabilized phase, calculate average values of the physical parameters on the part of the sequence of n-tuples (xi exec --- xn exec; Yexec norm) corresponding to the stabilized phase, so as to obtain a n- uplet ...., 2; 37) defining a recurrent point of said flight of the aircraft (2), and transmission to interface means (33).
    [0002]
    2. The method according to claim 1, wherein step (a) comprises separating the values (xi_exxxxec) of the exogenous parameters and values (vexec, 1 of the endogenous parameters.
    [0003]
    3. Method according to one of claims 1 and 2, wherein step (b) comprises, for each n-tup exec (xn exec; Yexec) of the received sequence, a preliminary step (b0) determination. an exogenous class of the tuple according to the values (xi_ exec xn exec) of the exogenous parameters of said tuple, among a plurality of exogenous classes each defined by the values (./C1, Xm) tElitpfl of the exogenous parameters d a subset of said set of n-tuplet reference sequences (x1, ... Xm; yi) / Elitm.
    [0004]
    4. The method of claim 3, wherein step (b0) comprises projecting said tuple in the context classes so as to identify the closest exogenous class according to a distance criterion.
    [0005]
    5. Method according to one of claims 3 and 4, wherein the regression model used in step (b) for a tuple (xl exec --- xn exec; Yexec) of the received sequence is the model. associated with the exogenous class determined for said tuple (x1 exec ... Xn exec; Yexec) -
    [0006]
    The method according to claim 5, comprising a prior phase of processing said set of reference sequences of n-tuples xni; database, comprising the implementation by data processing means (31) of steps of: (a0) classifying reference sequences of n-tuples --- xni; Yi) iE [[1, p]] so as to generate said plurality of exogenous classes; (ai) for each exogenous class, determining said regression model associated with the subset of said set of reference sequences of n-tuples xni; yi) ielitpil by a regression modeling the value y of the endogenous parameter as a function of the values x1 ... xn of the exogenous parameters from the set of n-tuples of the exogenous class; (a2) for each tuple ... xni; of the set of reference sequences, calculation for the exogenous class (j) of the n-tuple --- xni; Yi) iE [[1, p]] of an estimated value (9ii) of the endogenous parameter and an associated residue (resii); (a3) For each tuple ... xni; of the set of reference sequences, calculating the normalized value (yiinorm) of the endogenous parameter with respect to the exogenous parameters, so as to obtain a set of reference sequences of standardized n-tuples --- xni; Yi norm) iE [[1, p]]; (a4) for each reference sequence of normalized n-tuples (x1i-xni; Yi norm) iE [[1, p]]; identifying at least one stabilized phase in said normalized n-tuplet reference sequence (x1i-xni; Yi norm) iE [[1, p]]; a stabilized phase corresponding to a portion of said sequence representative of a flight time greater than a given threshold and in which the values of the standardized n-tuples --- xni; Yi norm) iE [[1, p]] are constant at a given variance; (a5) for each stabilized phase of each reference sequence, calculating mean values of the physical parameters on the portion of the sequence of n-tuples (xli ... xni; v norm) iE [[1, p]] corresponding to the stabilized phase, so as to obtain a tuple defining a potential recurrent point of a flight of the aircraft (2); (a6) classifying n-tuples defining a potential recurrent point obtained to generate a plurality of cells, each associated with a subset of the n-tuples defining a potential recurring point; (a7) for at least one generated cell, calculating a n-tuple (x1, ... xns; ys) of values of the physical parameters and a n-tuple (vars (xi) var (x); vars (y )) associated variance values so as to define a phase signature associated with the cell, and storing phase signatures in said data storage means database (32).
    [0007]
    7. The method according to claim 6, wherein the classification of steps (a0) and (a6) is implemented in accordance with a non-supervised method chosen from the k-means algorithm, the self-adaptive map method of Kohonen and hierarchical ascending classification (CAH).
    [0008]
    8. Method according to one of claims 6 and 7, wherein step (a6) comprises, for each cell generated, the calculation of a density of the cell defined as the number of reference sequences for which the cell comprises at least one tuple defining a potential recurrent point obtained, step (a7) being implemented for the cells having the highest density.
    [0009]
    9. Method according to one of claims 1 to 8, wherein each endogenous parameter is selected from a pressure output of a booster engine (1), a static pressure at the inlet of a combustion chamber of the engine (1). ), a temperature at the output of the engine booster (1), a temperature of the engine exhaust gas (1), a mass fuel flow rate at the inlet of a high-pressure compressor of the engine (1), and a speed high-pressure rotation at the inlet of the high-pressure compressor of the engine (1).
    [0010]
    10. Method according to one of claims 1 to 9, wherein each exogenous parameter is selected from an altitude, an inlet temperature of a motor blower (1), and a low pressure rotation speed at the inlet of the blower of the engine (1).
    [0011]
    11. Method according to one of claims 1 to 10, comprising adding the received sequence of n-tuples (x1 ex 'exec; exec) to said set of reference sequences of n-tuples (x1, ... xm). ;
    [0012]
    12. Equipment (3) for monitoring an aircraft engine (1) operating in a given environment, comprising: data processing means (31); data storage means (32); ) storing in a database: o a set of reference sequences of n-tuples --- xni; jiE [[1, p]] of physical parameter values relating to said aircraft engine (1), including at least one endogenous parameter specific to the operation of the engine (1) and at least one exogenous parameter specific to said environment, and o a set of phase signatures, each signature being defined by a n-tuple (x1, ... xns; ys) of values of the physical parameters and a tuple (v ar, (xv ar, (x 72); vars (y)) of associated variance values, - interface means (33), the equipment (3) being characterized in that the data processing means (31) are configured to implement: - A module for receiving a sequence of n-tuples (xl exec --- xn exec; Yexec) of values of the physical parameters relating to said aircraft engine (1) (2), said values being measured over time by means of sensors (20) such that each tuple of the sequence defines a point of a flight of said aircraft (2); - a calculation module, for each tuple exec ... Xn exec Yexec) of the received sequence, according to a regression model associated with a subset of said set of n-tuplet reference sequences (x1, ... xni; , a normalized value (Yexec norm) of the endogenous parameter with respect to the exogenous parameters, so as to obtain a sequence of standardized n-tuples (xl exec --- xn exec; Yexec norm); An identification module of at least one stabilized phase in said normalized sequence of n-tuples (x1 exec Xn exec; Yexec norm) from said set of phase signatures, a stabilized phase corresponding to a part of said representative sequence a flight time greater than a given threshold and in which the values of the standardized n-tuples (xi_ exec Xn exec; Yexec norm) coincide with the values of the tuple (x1, ... xns; ys) of a signature with the associated variance; A calculation module, for each stabilized phase, of average values of the physical parameters on the part of the sequence of n-tuples (x1 exec Xn exec; Yexec norm) corresponding to the stabilized phase, so as to obtain a tuple; ...., 2; 37) defining a recurrent point of said flight of the aircraft (2); and- a module for transmitting recurrent points to the interface means (33).
    [0013]
    Equipment according to claim 12, wherein the data processing module (31) is further configured to implement: a module for classifying reference sequences of n-tuples --- xni; Yi) iEL [1, p]] so as to generate a plurality of exogenous classes each defined by the values (xli Xni) ielitm of the exogenous parameters of a subset of said set of n-tuplet reference sequences Xni; ; A module for determining, for each exogenous class, said regression model associated with the subset of said set of reference sequences of n-tuples xni; yi) ielitpil by a regression modeling the value y of the endogenous parameter as a function of the values x1 ... xn of the exogenous parameters from the set of n-tuples of the exogenous class; - a calculation module, for each tuple xni; yi) ielitpil of the set of reference sequences, an estimated value (9ii) of the endogenous parameter and an associated residue (resii) for the exogenous class (j) of the tuple ... Xni; ; - a calculation module, for each tuple xni; yi) ielitpil of the set of reference sequences, of the normalized value (Yi] norm) of the endogenous parameter with respect to the exogenous parameters, so as to obtain a set of standardized reference sequences of n-tuples (xli ... xni; yi n01.77) iE1113) 11; An identification module, for each reference sequence of normalized tuples (xli ... xni; v norm) iE [[1, p]], of at least one stabilized phase in said reference sequence of n-tuples normalized --- xni; Yi norm) iE [[1, p]] a stabilized phase corresponding to part of said sequence representative of a flight time greater than a given threshold and in which the values of the standardized n-tuples (Xli --- Xni; Yi norm) .E [[1, p]] are constant at a given variance; A calculation module, for each stabilized phase of each reference sequence, of average values of the physical parameters on the part of the sequence of n-tuples (x1 / ... Xru; d17 / norm), EL [tp] corresponding to the stabilized phase, so as to obtain a tuple defining a potential recurrent point of a flight of the aircraft (2); A n-tuplet classification module defining a potential recurrent point obtained so as to generate a plurality of cells, each associated with a subset of the n-tuples 37) defining a potential recurring point; A calculation module, for at least one generated cell, of a n-tuple (x1, ... xns; ys) of values of the physical parameters and an n-tuple (vars (xi) vars (xn); (y)) associated variance values so as to define a phase signature associated with the cell, and storing phase signatures in said database of the data storage means (32).
    类似技术:
    公开号 | 公开日 | 专利标题
    FR3028331A1|2016-05-13|METHOD FOR MONITORING AN AIRCRAFT ENGINE IN OPERATION IN A GIVEN ENVIRONMENT
    EP2368161B1|2016-11-09|Detection of anomalies in an aircraft engine
    EP2376988B1|2013-07-10|Standardisation of data used for monitoring an aircraft engine
    CA2746543C|2018-01-02|Identification of defects in an aircraft engine
    EP2912526B1|2017-12-06|System for monitoring a set of components of a device
    FR2963448A1|2012-02-03|METHOD AND SYSTEM FOR ANALYSIS OF FLIGHT DATA RECORDED DURING THE FLIGHT OF AN AIRCRAFT.
    FR2965915A1|2012-04-13|SYSTEM FOR MONITORING AN AIRCRAFT ENGINE TEST BENCH
    EP3039497B1|2018-08-29|Monitoring of an aircraft engine to anticipate maintenance operations
    FR3046268A1|2017-06-30|AIRCRAFT FLIGHT DATA OPERATION SYSTEM
    EP2072399A1|2009-06-24|Verfahren und System zur Charakterisierung und Zählung von Überschreitungen eines Schwellwerts durch einen Funktionsparameter eines Flugzeugmotors
    FR3035232A1|2016-10-21|SYSTEM FOR MONITORING THE HEALTH CONDITION OF AN ENGINE AND ASSOCIATED CONFIGURATION METHOD
    WO2016087770A1|2016-06-09|Method for monitoring the operation of a turbomachine
    EP3871078A1|2021-09-01|Computing environment system for monitoring aircraft engines
    FR3010200A1|2015-03-06|METHOD AND DEVICE FOR NORMALIZING VALUES OF OPERATING PARAMETERS OF AN AIRCRAFT ENGINE
    FR2970796A1|2012-07-27|Method for processing sequence of failure messages for discriminating failure system in aircraft, involves application of pattern matching process to data obtained from test to recognize and classify data using knowledge model
    EP3846047A1|2021-07-07|Method and system for identifying relevant variables
    FR3076634A1|2019-07-12|METHOD FOR ANALYZING PLATFORM FAILURE SYMPTOMS, AND SYSTEM THEREOF
    FR3076632A1|2019-07-12|METHOD FOR ANALYZING MAINTENANCE MESSAGES GENERATED BY AT LEAST ONE PLATFORM, AND ASSOCIATED SYSTEM
    WO2020229778A1|2020-11-19|Model resetting in a turbine engine
    WO2021069824A1|2021-04-15|Apparatus, method and computer program for monitoring an aircraft engine
    WO2020217004A1|2020-10-29|System and method for monitoring an aircraft engine
    FR3096137A1|2020-11-20|PS3 model registration in a turbomachine
    FR3096136A1|2020-11-20|Analysis of the aging of a turbomachine
    CA3041415A1|2018-05-03|Method and system for monitoring the soundness of helicopters
    同族专利:
    公开号 | 公开日
    WO2016075409A1|2016-05-19|
    FR3028331B1|2016-12-30|
    GB2558017A|2018-07-04|
    GB2558017B|2020-12-16|
    GB201709066D0|2017-07-19|
    US20180170580A1|2018-06-21|
    US10486833B2|2019-11-26|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    EP1630633A2|2004-08-26|2006-03-01|United Technologies Corporation|System for gas turbine health monitoring data fusion|
    FR2939928A1|2008-12-15|2010-06-18|Snecma|STANDARDIZATION OF DATA USED FOR MONITORING AN AIRCRAFT ENGINE|
    FR2971595A1|2011-02-15|2012-08-17|Snecma|MONITORING AN AIRCRAFT ENGINE TO ANTICIPATE MAINTENANCE OPERATIONS|FR3074590A1|2017-12-04|2019-06-07|Societe Air France|METHOD FOR PREDICTING AN OPERATING FAILURE OF ONE OR MORE EQUIPMENTS OF AN ASSEMBLY|
    FR3111200A1|2020-06-08|2021-12-10|Airbus Helicopters|Method and system for controlling a level of damage to at least one part of an aircraft, associated aircraft.|US4215412A|1978-07-13|1980-07-29|The Boeing Company|Real time performance monitoring of gas turbine engines|
    US4729102A|1984-10-24|1988-03-01|Sundstrand Data Control, Inc.|Aircraft data acquisition and recording system|
    US6236955B1|1998-07-31|2001-05-22|Gary J. Summers|Management training simulation method and system|
    US7440906B1|2001-09-04|2008-10-21|Accenture Global Services Gmbh|Identification, categorization, and integration of unplanned maintenance, repair and overhaul work on mechanical equipment|
    EP1570424A2|2002-10-01|2005-09-07|Target Discovery, Inc.|Methods and compositions utilizing evolutionary computation techniques and differential data sets|
    US7005986B2|2003-08-19|2006-02-28|Kardios Corporation|Remote temperature monitoring apparatus|
    US20150235143A1|2003-12-30|2015-08-20|Kantrack Llc|Transfer Learning For Predictive Model Development|
    US7487029B2|2004-05-21|2009-02-03|Pratt & Whitney Canada|Method of monitoring gas turbine engine operation|
    FR2965915B1|2010-10-11|2013-08-02|Snecma|SYSTEM FOR MONITORING AN AIRCRAFT ENGINE TEST BENCH|
    US8775341B1|2010-10-26|2014-07-08|Michael Lamport Commons|Intelligent control with hierarchical stacked neural networks|
    US8689173B2|2011-01-21|2014-04-01|International Business Machines Corporation|Detecting design patterns in models by utilizing transformation language|
    US9916538B2|2012-09-15|2018-03-13|Z Advanced Computing, Inc.|Method and system for feature detection|
    US9958867B2|2012-01-13|2018-05-01|Borealis Technical Limited|Monitoring and control system for enhancing ground movement safety in aircraft equipped with non-engine drive means|
    US9466024B2|2013-03-15|2016-10-11|Northrop Grumman Systems Corporation|Learning health systems and methods|
    FR3006785B1|2013-06-10|2015-07-03|Snecma|METHODS OF CREATING A DATABASE AND DEVELOPING A CARD OF OPERATING STATES OF AIRCRAFT ENGINES, AND A METHOD OF MONITORING THE OPERATION OF AN ASSOCIATED AIRCRAFT ENGINE|
    EP3008528B1|2013-06-14|2020-02-26|Wallace E. Larimore|A method and system of dynamic model identification for monitoring and control of dynamic machines with variable structure or variable operation conditions|
    FR3026785B1|2014-10-06|2019-08-02|Safran Aircraft Engines|MONITORING A SET OF THE PROPULSIVE SYSTEM OF AN AIRCRAFT|US11267587B2|2019-06-10|2022-03-08|The Boeing Company|Methods and systems for an ad hoc network sensor system|
    FR3101669A1|2019-10-07|2021-04-09|Safran|Aircraft engine monitoring device, method and computer program|
    WO2021175493A1|2020-03-04|2021-09-10|Nuovo Pignone Tecnologie - S.R.L.|Hybrid risk model for maintenance optimization and system for executing such method|
    CN111907730B|2020-07-31|2021-12-10|西安电子科技大学|Real-time online unmanned aerial vehicle multi-fault abnormity detection method and device|
    CN111829425A|2020-08-06|2020-10-27|厦门航空有限公司|Health monitoring method and system for civil aircraft leading edge position sensor|
    法律状态:
    2015-11-16| PLFP| Fee payment|Year of fee payment: 2 |
    2016-05-13| PLSC| Publication of the preliminary search report|Effective date: 20160513 |
    2016-11-09| PLFP| Fee payment|Year of fee payment: 3 |
    2017-10-20| PLFP| Fee payment|Year of fee payment: 4 |
    2018-02-09| CD| Change of name or company name|Owner name: SAFRAN AIRCRAFT ENGINES, FR Effective date: 20170717 |
    2018-10-24| PLFP| Fee payment|Year of fee payment: 5 |
    2019-10-22| PLFP| Fee payment|Year of fee payment: 6 |
    2020-10-21| PLFP| Fee payment|Year of fee payment: 7 |
    2021-10-20| PLFP| Fee payment|Year of fee payment: 8 |
    优先权:
    申请号 | 申请日 | 专利标题
    FR1460853A|FR3028331B1|2014-11-10|2014-11-10|METHOD FOR MONITORING AN AIRCRAFT ENGINE IN OPERATION IN A GIVEN ENVIRONMENT|FR1460853A| FR3028331B1|2014-11-10|2014-11-10|METHOD FOR MONITORING AN AIRCRAFT ENGINE IN OPERATION IN A GIVEN ENVIRONMENT|
    US15/525,574| US10486833B2|2014-11-10|2015-11-10|Method for monitoring an aircraft engine operating in a given environment|
    GB1709066.3A| GB2558017B|2014-11-10|2015-11-10|Method for monitoring an aircraft engine operating in a given environment|
    PCT/FR2015/053050| WO2016075409A1|2014-11-10|2015-11-10|Method for monitoring an aircraft engine operating in a given environment|
    [返回顶部]