• 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 invention relates to a method for detecting electric arcs for DC-source electrical installations, implemented by a detection device comprising a calculation processor, arranged between the DC power source and a load. The method of the invention comprises obtaining (40, 42) a digitized time signal from the electric current supplied by the DC power source, and, for a plurality of successive time windows of predetermined time duration, an extraction ( 44) of a digitized time sub-signal of said digitized time signal corresponding to a current time window, followed by an application (46) of band-pass filtering to said digitized time sub-signal. The method also includes calculating (50, 52) a fourth order statistical value of the filtered digitized time sub-signal, and comparing (54) the fourth order statistical value calculated at a first statistical threshold and / or at a second statistical threshold, and in the case where the fourth-order statistical value is less than the first statistical threshold, or in the case where the fourth-order statistical value is greater than the second statistical threshold, storing an indication positive detection for the current time window. An arcing detection (60) is performed when the number of stored positive detection indications exceeds a predetermined detection threshold.
    公开号:FR3023377A1
    申请号:FR1456477
    申请日:2014-07-04
    公开日:2016-01-08
    发明作者:Mounir Djeddi
    申请人:Schneider Electric Industries SAS；
    IPC主号:
    专利说明:

    [0001] The present invention relates to a method and device for the detection of electric arcs for DC-source electrical installations.
    [0002] It is in the field of protection of electrical installations. The invention finds particular application in the field of photovoltaic installations, comprising photovoltaic panels as sources of DC electrical energy. In photovoltaic installations, an inverter device capable of transforming the direct electric current into alternating current to be injected into the electrical distribution network is connected to the photovoltaic panels. In general, the invention applies to any other type of electrical installation comprising one or more sources of direct electric current, connected to a load, particularly when the connection cable between current sources and loads is long.
    [0003] In such installations, it is possible to observe, over the length of the connection cable, insulation faults. Electric arcs can then occur, inducing a risk of fire and a safety risk for maintenance agents. In addition, downstream loads can also be damaged, resulting in repair and replacement costs.
    [0004] The problems related to the appearance of arcing are well known, and there are various methods of detection of arcing as early as possible, in order to achieve a break in the electrical circuit and prevent the occurrence of fire , the accident of maintenance agent and the damage of the loads. In the field of direct current sources for photovoltaic installations, it is particularly difficult to apply certain known methods, due to the presence of resistive loads and the resistance of the long connection cable. In some cases the arc current is low and it has been observed by simulation that the spectrum of the current in case of presence of arc is very close to the spectrum of the current in the absence of arc, which makes the electric arc detection very difficult.
    [0005] It is an object of the invention to provide an electric arc detection method for DC source electrical installations, which is effective even in cases where the electric arc current is low. For this purpose, the invention proposes, according to a first aspect, a method for detecting electric arcs for DC-source electrical installations, implemented by a detection device comprising a calculation processor, arranged between the source of DC current and a load.
    [0006] The method of the invention comprises steps of: obtaining a digitized temporal signal from the electric current supplied by the direct current source, for a plurality of successive time windows of predetermined temporal duration: extraction of a digitized time sub-signal of said digitized time signal corresponding to a current time window, -application of a band-pass filtering to said digitized time sub-signal, -calculation of a fourth-order statistical value of the filtered digitized time sub-signal comparing the four-fold statistical value calculated at a first statistical threshold and / or a second statistical threshold, and in the case where the fourth-order statistical value is less than the first statistical threshold, or in the case where the fourth order statistical value is greater than the second statistical threshold, storing a positive detection indication for the time window yard ante, arcing presence detection when the number of stored positive detection indications exceeds a predetermined detection threshold. Advantageously, the electric arc detection method according to the invention comprises the extraction of temporal sub-signals and the analysis of fourth-order statistics in several time windows to detect the presence of an electric arc. The method according to the invention is effective even when the arc current is low, and it fast from a computational point of view. The arcing detection method according to the invention may also have one or more of the features below. This method comprises, for a said digitized time sub-signal corresponding to a current time window, a determination of an envelope maximum value of the digitized time sub-signal, and an update of said positive detection indication according to a current time window. the calculated fourth order statistical value and the calculated envelope maximum value. The determination of an envelope maximum value of the digitized time sub-signal is performed when said fourth-order statistical value is greater than the first statistical threshold and lower than the second statistical threshold. The method further comprises, when the fourth order statistical value is greater than the first statistical threshold and less than an intermediate statistical threshold, a comparison of the envelope maximum value with a first envelope threshold, and a detection of arcing presence if the envelope maximum value is greater than said first envelope threshold. When the fourth order statistical value is greater than the intermediate statistical threshold and lower than the second statistical threshold, the method comprises a comparison of the envelope maximum value with a second envelope threshold, and an arc presence detection. if the envelope maximum value is greater than said second envelope threshold. The method comprises calculating a coefficient of variation depending on the average and the second order statistical value of the digitized time sub-signal corresponding to the current time window before bandpass filtering, and an update of said indication. positive detection when said coefficient of variation exceeds a predetermined variation threshold. For each current time window, the storage of an electric arc detection indication in a circular buffer of a predetermined size is implemented. According to one characteristic, the detection threshold is calculated according to the size of the circular buffer. According to one characteristic, two successive time windows have an overlap of a predetermined recovery time.
    [0007] The bandpass filtering is defined by a first frequency F1 and a second frequency F2, said first frequency being a function of a cutoff frequency of said load, said second frequency being a function of a sampling frequency of an analog converter digital device for obtaining said digitized time signal.
    [0008] According to a second aspect, the invention relates to an electric arc detection device for DC-source electrical installations, comprising a calculation processor, arranged between the DC power source and a load. The device comprises: - a module for obtaining a time signal digitized from the electric current supplied by the DC power source, for application to a plurality of successive time windows of predetermined time duration, - an extraction unit a digitized time sub-signal of said digitized time signal corresponding to a current time window; a band-pass filtering module to said digitized time sub-signal; a calculation module of a fourth-order statistical value of the sub-time signal; filtered digitized temporal signal, - a decision module, able to perform a comparison of the calculated fourth order statistical value with a first statistical threshold and / or a second statistical threshold, and in the case where the statistical value of fourth order is less than the first statistical threshold, or in the case where the fourth order statistical value is greater than the second statistical threshold, a providing a positive detection indication for the current time window; - the decision module being able to detect presence of an electric arc when the number of stored positive detection indications exceeds a predetermined detection threshold. The arcing device according to the invention may also have one or more of the features below. It comprises a module triggering a security unit in response to a trigger signal sent by the decision module. The unit for extracting a digitized time sub-signal from said digitized time signal corresponding to a current time window comprises an electric current sensor and an analog-digital converter having a given sampling frequency.
    [0009] The band-pass filtering module is the band-pass filtering is able to perform a filtering between a first frequency F1 and a second frequency F2, said first frequency being a function of a cutoff frequency of said load, said second frequency being function the sampling frequency. It comprises an envelope calculation module of said digitized time sub-signal corresponding to a current time window, an envelope maximum value of the digitized time sub-signal being transmitted to said decision module. It comprises a variation calculation module able to calculate a coefficient of variation depending on the average and the second order statistical value of the digitized temporal sub-signal corresponding to the current time window before bandpass filtering, and to transmit the coefficient of variation calculated at said decision module. Other characteristics and advantages of the invention will emerge from the description which is given below, by way of indication and in no way limitative, with reference to the appended figures, among which: FIG. 1 schematically represents an electrical installation comprising a device for arcing detection according to one embodiment of the invention; FIG. 2 diagrammatically represents a set of functional blocks of an arcing detection device according to one embodiment of the invention; FIG. 3 illustrates functional modules implemented by a device for calculating an arcing detection device according to the invention; FIG. 4 schematically illustrates samples of digitized time sub-signals; FIG. 5 is a block diagram of the main steps of an arcing detection method according to one embodiment of the invention.
    [0010] Figure 1 schematically illustrates an electrical installation 2 in which the invention finds an application. In the illustrated embodiment it is a photovoltaic electrical installation 2, comprising a DC source 4, formed for example of a set of solar panels, connected in the circuit of the electrical installation 2 to an inverter 6. The inverter 6 is capable of transforming the direct electric current supplied by the source 4 into alternating electric current of frequency 50 Hz, making it possible to supply an electrical network 8. The circuit of the installation 2 also comprises a device for detecting presence of electric arcs 10, the main functional blocks of which are described in more detail with reference to FIG. 2. The arcing detection device 10 comprises an electric current sensor 12 which is able to recover the corresponding signal. to the electric current DC, and to digitize it to provide a digitized time signal at the input of a computing device 14 able to perform digital signal processing. The computing device 14 is for example a microcontroller or a digital signal processor, also called DSP for "Digital Signal Processor". The device 14 comprises, in particular, a microprocessor, memories and input-output interfaces 16 making it possible to send control commands to a security unit 18, able to open the circuit in such a way as to eliminate the arc propagation. electrical detected. The device 14 is programmed to implement decision functional modules as schematically illustrated in FIG. 3. As illustrated in FIG. 3, the DC electric current sensor 12 retrieves the corresponding signal, and the conversion unit 20 analog / digital provides a digitized time signal S which is stored as it is acquired, preferably in a buffer memory (or buffer in English terminology) of the digital samples. A digitized temporal sub-signal extraction unit 22 corresponding to a sliding time window, defined between a time instant To and a time instant To + At, is applied to extract a digitized time sub-signal SF. For example, At = 0.2 seconds, which corresponds to K samples of the digital signal. The extraction unit 22 is applied successively by applying the sliding window, two successive extractions preferably having a recovery zone of dt seconds corresponding to r samples.
    [0011] For example, the overlap zone is equal to dt = 0.5x At. More generally, dt is between 0.1 x At and 0.9 x. In FIG. 4, the samples S, of the digitized time signal are stored in the memory 21, and digitized time sub-signals Swi, Sw2 and so on are extracted from the buffer memory 21, for time windows W1, W2 with an overlap of r = 2 samples in this example. The digitized time sub-signals are then processed by the respective processing modules 24 and 26 to 28. The filtering module 24 applies a band-pass filtering between a first frequency F1 and a second frequency F2, for example F1 = 50 kHz ( kilo-Hertz) and F2 = 90kHz. Other first frequency and second frequency values may be selected for the definition of bandpass filtering. More generally, the first frequency is chosen as a function of the cutoff frequency of the load 6, and the second frequency is chosen as a function of the sampling frequency of the digital analog conversion applied to the time signal representative of the current electric. For example, for sampling at 200kHz, a second frequency of 90kHz is selected. In a first embodiment, the filtering module 24 is implemented by the computing device 14 as described above.
    [0012] In an alternative embodiment, the filtering module 24 is implemented by an analog filtering circuit. A filtered digitized time sub-signal is obtained at the output of the filtering module 24. The statistical calculation module 26 performs a calculation of a fourth order statistical value or kurtosis, also called flattening coefficient, of each sub-signal filtered digital time filter, and the envelope calculation module 28 calculates the envelope maximum of each filtered digitized time sub-signal, as explained in more detail below. The variation calculation module 30 performs a calculation of the coefficient of variation of the time sub-signal before bandpass filtering, for each digitized time sub-signal extracted by the extraction unit 22. In one embodiment, the modules Statistical calculation 26, envelope calculation 28 and variation calculation 30 are carried out substantially in parallel. Each of the calculation modules provides an arcing presence detection indicator, if any, for each digitized time sub-signal corresponding to a time window, to a decision module 32. In an alternative embodiment, the module Statistical calculation 26 is applied in a first step, and the envelope calculation modules 28 and variation calculation modules 30 are subsequently applied, depending on the result of the statistical calculation module 26.
    [0013] The decision module 32 stores arc presence detection indications for a plurality of digitized time sub-signals each corresponding to a time window, and performs an electric arc detection determination according to the plurality of the detection indications. arc. In one embodiment, the arc detection indications are represented in the form of flags taking a binary value respectively indicating the detection of the presence or absence of an electric arc for a digitized temporal sub-signal corresponding to a given time window. , and are stored in an arc detection buffer, of circular type and of a given capacity N. The decision module 32 decides the presence of an electric arc when the arc detection buffer memory contains at least n positive indications among the N stored indications. Preferably, n is between 30% of N and 100% of N. In the event of a positive decision by the decision module 32, this module sends a trigger signal to a module 34 for triggering the security unit 18, to cause an opening of the electrical circuit. FIG. 5 is a block diagram of the main steps of an electric arc detection method for DC-source electrical installations according to one embodiment of the invention. In a first electrical signal acquisition step 40, the time signal representative of the direct electric current from the DC source is acquired.
    [0014] This time signal is digitized by an analog digital converter during the scanning step 42, and the digitized time signal samples obtained are stored in a buffer of the digital samples. In an exemplary embodiment, a scanning at 200 kHz is performed.
    [0015] A digitized time sub-signal is extracted during the sub-signal extraction step 44, for a current time window between an initial time To and a given final time To + At. The digitized time signal is filtered during a band-pass filtering step 46, for example by a 6-order Butterworth filter between a first frequency of 50 kHz and a second frequency of 90 kHz. The set of digital samples of the extracted digitized time sub-signal is also provided at the input of a variation coefficient calculation step 48. For a digitized time sub-signal Sw composed of samples Sw (1) to Sw (K ), the coefficient of variation C, (Sw) is given by: C, (Sw) = 100x Sw o- (Sw) (Eq 1) Where, u (Sw) is the mean of the digitized time sub-signal Sw, or first-order statistic:, u (Sw) = E [Sw 1 = -K 1 Sw (i) (Eq 2) E [X] denoting the expected value of a variable X.
    [0016] And 6 (Sw) is the standard deviation of the digitized temporal sub-signal Sw, or second-order statistic: 6 (Sw) = VE [(Sw -, u (Sw)) 2 = (Sw (i) ii ( Sw)) 2 (Eq 3) After filtering the digitized time sub-signal during the bandpass filtering step 46, the average and the variance of the filtered digitized time sub-signal SWF are calculated in the calculation step 50 1st and 2nd order statistics after filtering:, u (SwF) = E [SWF = 1 SwF (i) (Eq 4) K 62 (SwF) = EkSwF ii (SwF)) 21 (Eq 5) The statistical value of order 4 or kurtosis of the filtered digitized time sub-signal is calculated in the higher order statistic calculation step 52, performed SWF after the calculation of the first and second order statistics.
    [0017] Kurtosis is defined for a SWF filtered digitized temporal sub-signal by: EkSwF -, u (SwF)) 4] K (SwF) (74 (SWF (Eq 6)) For a given filtered SWF filtered temporal sub-signal This results in a kurtosis value: When the distribution of the samples of the SWF filtered digitized sub-signal is normalized Gaussian, the kurtosis value is K = 3. The kurtosis value obtained for a filtered SWF filtered temporal sub-signal. is provided at a detection step 54 for a given time window The detection step 54 implements arc detection tests each providing a presence or absence of electric arc detection indicator, for the sub a digitized temporal signal corresponding to the current time window It has been found, following experimental tests carried out by the inventors, that when the value of kurtosis K (SWF) calculated for a given SWF filtered digital time sub-signal is less than a first statistical threshold K1, or is greater than a second statistical threshold K2, the detection of the presence of an electric arc is positive.
    [0018] Preferably, the first statistical threshold is K1 = 2.1 and the second statistical threshold is K2 = 5. In a general way K1 = 2.1 ± AK, and K2 = 5 ± AK2 here AK, and AK2 are defined so that the detection of the arc is always optimal, with AK, = 0.1 and AK2 = 0 1. A first test performed in the detection step 54 consists in comparing the value of the kurtosis K (SWF) calculated at the first threshold K1 and the second threshold K2. If the first test indicates that K (SWF) Ki or K (SWF) K2, a first indi binary indicator is then set to 1. More generally, a first indicator Indi is set to a value indicating an arc detection positive electric current for the digitized time sub-signal Sw considered. The method also comprises a step 56 for calculating an envelope signal of the digitized time sub-signal Sw considered and a step 58 for calculating the envelope maximum of this sub-signal. Step 56 is performed substantially in parallel with the step of calculating higher order statistics 52 or after the step of calculating higher order statistics 52. It comprises the calculation of an envelope signal from of the filtered digitized temporal sub-signal SWF Y (n) = SWF (i) J. II (S wF (i)) (Eq 7) Where j is the unit of the imaginary domain and H (X) is the Hilbert transform a signal X. Then, in step 58, the maximum absolute value of the analytical signal Y (i) is calculated.
    [0019] The envelope maximum, denoted MaxEnv (SwF), is calculated as follows: MaxEnv (SwF) = MAX (Y (i)) (Eq 8) Where MAX (-) is the maximum function and 117 (i) 1 is the absolute value of the amplitude of the signal Y (i). The envelope maximum value obtained is also provided at the detection step 54. In the detection step 54, the envelope maximum value MaxEnv (SwF) is used with the value of kurtosis K (SwF) calculated for the current filtered digitized temporal sub-signal to perform a second arc presence detection test. If the value of K (Swf) is between the first statistical threshold K1 is an intermediate statistical threshold value Ka and if the maximum envelope value of the sub-MaxEnv signal (Swf) is greater than a first envelope threshold ME , then a second indicator India is set to a value indicating a positive electric arc detection for the digitized sub-time signal Sw considered. Preferably Kint = 3.3 and ME1 = 0.3 If the value of K (Swf) is between the intermediate statistical threshold value Ka and the second statistical threshold K2, and if the maximum envelope value MaxEnv (Swf) ) is greater than a second envelope threshold ME2, while the second indicator India is set to a value indicating a positive electric arc detection for the digitized sub-time signal Sw considered.
    [0020] Preferably Ka = 3.3 and ME2 = 0.03 It will also be noted that K2> K1. Finally, in the detection step 54, the value of the coefficient of variation calculated during the variation coefficient calculation step 48 is used to perform a third arcing presence detection test.
    [0021] The value of the calculated coefficient of variation is compared with a variation threshold value V, for example equal to 15, and preferably between 15% and 100%. If the threshold of variation is exceeded, so if Cv (Sw)> V, then a third indicator Ind3 is set to a value indicating a positive electric arc detection for the sub-signal SwF considered.
    [0022] Finally, during the detection step 54, several values calculated for the digitized time sub-signal corresponding to the current time window are compared with thresholds, and three electric arc presence detection indicators are updated.
    [0023] The three indicators provide an indication of electric arc detection, positive or negative, for the current time window. If one of the three updated indicators Incl, India and Ind3 is positive, therefore indicates positive detection of arcing presence, then the arcing detection indication for the current time window is positive.
    [0024] If all three Incli, India, and Ind3 indicators are negative, then the arc detection indication for the current time window is negative. If the electric arc detection indication for the current time window is positive, then the detection step 54 is followed by the arc decision step 60 based on the tests performed for a plurality of subsets. digitized temporal signals extracted. According to one embodiment, the electric arc detection indication for the current time window is updated in a circular buffer, called an arc detection buffer, for storing detection test results for a given time. predetermined number N of digitized time sub-signals, extracted successively using a sliding time window in the predetermined duration time AT. In practice, the positive electric arc detection indication for the current time window is achieved by setting a corresponding bit in the arc detection buffer memory, whereas a negative detection indication is translated. by setting the corresponding bit to 0. For example, the size N of the arc detection buffer is given by the formula: N = J (Eq 9) AT - dt Where 0.1 is the numerical rounding of the value x, fs is the sampling frequency , AT is the time window duration and the recovery time between two successive time windows. It is determined, in the arc decision decision step 60, whether the number n of positive detection indications by digitized time sub-signal is greater than or equal to a detection threshold calculated as a function of N, by example 0.3xN.
    [0025] In general, a coefficient P representative of a rate of positive detection indications among the N last arc presence detections made is used and it is checked whether n PxN. In practice, when the electric arc detection indication is a bit equal to 1, it is sufficient to sum all the values stored in the arc detection buffer memory to calculate the value n. When the decision is positive, a trigger signal of the security unit is sent to the security triggering step 62, to induce an opening of the circuit.
    [0026] In the case of a negative decision, step 60, as well as the detection step 54 in the event of negative detection for all the tests performed, is followed by step 44 of extracting samples from a sub-sample. digitized time signal for a new current time window, between To + dt and To + dt + AT Alternatively, only the first detection test based on the fourth order statistic calculated for each extracted digitized time sub-signal is used. This variant is faster from a computational point of view. In another variant, step 48 is performed only if the kurtosis value calculated for a digitized time sub-signal considered is between the first threshold K1 and the second threshold K2. Then, steps 56 and 58 are performed only if the kurtosis value calculated for a filtered digitized time sub-signal considered is between the first threshold K1 and the second threshold K2 and the value of the coefficient of variation is below the threshold. of variation. In this variant, the first, second and third detection tests are not all carried out systematically, they are performed separately and conditionally to the result of a previous test. Time and computational resources are saved in this variant embodiment. In the case where the three arc presence detection tests are carried out, it is possible to detect several types of electric arcs, by monitoring various characteristics of the electrical signal.
    权利要求:
    Claims (16)
    [0001]
    CLAIMS1.- A method for detecting electric arcs for DC-source electrical installations, implemented by a detection device comprising a calculation processor, arranged between the DC power source and a load, characterized in that comprises steps of: -obinding (40, 42) of a digitized time signal from the electric current supplied by the DC power source, -for a plurality of successive time windows of predetermined temporal duration: -extraction (44) d a digitized time sub-signal of said digitized time signal corresponding to a current time window; application of a band-pass filtering to said digitized time sub-signal; a calculation of a statistical value of fourth order of the digitized filtered temporal sub-signal, -comparison (54) of the fourth order statistical value calculated at a first statistical threshold and / or at a second statistical threshold ue, and in the case where the fourth-order statistical value is lower than the first statistical threshold, or in the case where the fourth-order statistical value is greater than the second statistical threshold, storing a positive detection indication for the current time window, arcing presence detection (60) when the number of stored positive detection indications exceeds a predetermined detection threshold. 25
    [0002]
    2. A method for detecting arcs according to claim 1, characterized in that it comprises, for a said digitized time sub-signal corresponding to a current time window, a determination (56, 58) of a value of envelope maximum of the digitized time sub-signal, and an update of said positive detection indication as a function of the calculated fourth order statistical value and the calculated envelope maximum value.
    [0003]
    3. An arcing detection method according to claim 2, characterized in that the determination of an envelope maximum value of the digitized time sub-signal is performed when said fourth-order statistical value is greater than first statistical threshold and below the second statistical threshold.
    [0004]
    4. A method for detecting electric arcs according to one of claims 2 or 3, characterized in that it comprises: when the statistical value of fourth order is greater than the first statistical threshold and less than an intermediate statistical threshold a comparison of the envelope maximum value at a first envelope threshold, and an arcing presence detection if the envelope maximum value is greater than said first envelope threshold.
    [0005]
    5. A method of arcing detection according to claim 4, characterized in that it comprises: when the fourth order statistical value is greater than the intermediate statistical threshold and lower than the second statistical threshold, a comparison of the value envelope maximum at a second envelope threshold, and an arcing presence detection if the envelope maximum value is greater than said second envelope threshold.
    [0006]
    6. A method of detecting electric arcs according to any one of claims 1 or 2, characterized in that it comprises a calculation (48) of a coefficient of variation depending on the mean and the statistical value of second order of the digitized time sub-signal corresponding to the current time window before bandpass filtering, and an update of said positive detection indication when said coefficient of variation exceeds a predetermined threshold of variation.
    [0007]
    7. A method for detecting electric arcs according to any one of claims 1 to 6, characterized in that it comprises, for each current time window, the storage of an electric arc detection indication in a memory circular buffer of predetermined size.
    [0008]
    8. A method of arcing detection according to claim 7, characterized in that said detection threshold is calculated according to the size of the circular buffer.
    [0009]
    9. A method of arcing detection according to any one of claims 1 to 8, characterized in that two successive time windows have a recovery of a predetermined recovery time.
    [0010]
    10. A method for detecting electric arcs according to any one of claims 1 to 9, characterized in that the band-pass filtering (46) is defined by a first frequency F1 and a second frequency F2, said first frequency being a function of a cut-off frequency of said load, said second frequency being a function of a sampling frequency of an analog-digital converter making it possible to obtain said digitized time signal.
    [0011]
    11. A device for detecting electric arcs for DC-source electrical installations, comprising a calculation processor, arranged between the DC power source and a load, characterized in that it comprises: a module for obtaining a digitized time signal from the electric current supplied by the DC source, for application to a plurality of successive time windows of predetermined time duration, -an extraction unit (22) of a temporal sub-signal digitized from said digitized time signal corresponding to a current time window; - a bandpass filtering module (24) to said digitized time sub-signal; - a module (26) for calculating a fourth-order statistical value of the sub-signal; filtered digitized temporal signal, - a decision module (32) able to perform a comparison of the calculated fourth order statistical value with a first statistical threshold and / or a second statistical threshold, and in the case where the fourth-order statistical value is lower than the first statistical threshold, or in the case where the fourth-order statistical value is greater than the second statistical threshold, storing a positive detection indication for the current time window, - the decision module (32) being able to detect the presence of an electric arc when the number of stored positive detection indications exceeds a predetermined detection threshold.
    [0012]
    12. An arcing detection device according to claim 11, characterized in that it comprises a tripping module (34) of a security unit in response to a trigger signal sent by the decision module (32). .
    [0013]
    13.- arcing detection device according to any one of claims 11 or 12, characterized in that the extraction unit (22) of a digitized sub-timed signal of said digitized temporal signal corresponding to a time window current includes an electric current sensor and an analog-to-digital converter having a given sampling frequency.
    [0014]
    14. An arcing detection device according to claim 13, characterized in that the bandpass filtering module (24) is the bandpass filtering is capable of performing a filtering between a first frequency F1 and a second frequency F2, said first frequency being a function of a cutoff frequency of said load, said second frequency being a function of the sampling frequency.
    [0015]
    15.- arcing detection device according to any one of claims 11 to 14, characterized in that it comprises a module (28) of envelope calculation of said digitized temporal sub-signal corresponding to a current time window , an envelope maximum value of the digitized time sub-signal being transmitted to said decision module (32).
    [0016]
    16.- arcing detection device according to any one of claims 11 to 15, characterized in that it comprises a module (30) for calculating variation capable of calculating a coefficient of variation depending on the average and of the second order statistical value of the digitized time sub-signal corresponding to the current time window before bandpass filtering, and transmitting the calculated coefficient of variation to said decision module (32).
    类似技术:
    公开号 | 公开日 | 专利标题
    EP2966461B1|2017-06-14|Method and device for detecting electrical arcs for electrical systems with a dc source
    EP1815260B1|2008-03-19|Method and device for detecting electric arc phenomenon on at least one electric cable
    EP3187842B1|2018-09-05|Method for detecting an electric arc by analysing the acoustic signature thereof
    EP2806518B1|2021-01-27|Serial arc detection based on harmonic content of DC current signal
    EP2788831B1|2019-01-09|Method and device for detecting electric arc in a photovoltaic installation
    EP2169799A2|2010-03-31|Directional detection of a ground fault
    EP2959304A1|2015-12-30|Method and device for detecting an electric arc in a photovoltaic facility
    EP1890165B1|2012-03-21|Method of directional detection of a fault in the ground connection and device for implementing the same
    EP3384592B1|2020-04-22|Method and device for detecting an electric arc in a photovoltaic installation
    EP3039439A1|2016-07-06|Electric arc detection in photovoltaic installations
    EP1845599A1|2007-10-17|Electric arc detection device, cut-off device comprising such a device and electric arc detection method
    FR2749986A1|1997-12-19|APPARATUS FOR DETECTING INSULATION FAILURE OF A CONNECTED DEVICE IN AN ELECTRICAL POWER DISTRIBUTION OR TRANSPORT NETWORK AND CORRESPONDING DETECTION METHOD
    EP2959303A1|2015-12-30|Method and device for detecting an electric arc in a photovoltaic installation
    EP3264116B1|2021-03-24|Device for detecting an electric arc from the acoustic signature thereof
    EP3384594B1|2019-09-11|Method and device for evaluating the energy produced by an electric arc in photovoltaic apparatus
    EP3384595B1|2019-09-11|Method and device for evaluating the energy produced by an electric arc in a photovoltaic installation
    EP3098830B1|2018-11-07|Device for monitoring partial discharges in an electrical network
    Sultana et al.2017|Novel control methodology for detecting series arc in dc circuits
    CN109768769B|2020-07-07|Fault arc identification method and device for photovoltaic system
    EP3920204A1|2021-12-08|Method and device for detecting the activation of a cut-off device, and electrical cabinet comprising such a device
    同族专利:
    公开号 | 公开日
    EP2966461A1|2016-01-13|
    EP2966461B1|2017-06-14|
    FR3023377B1|2016-08-12|
    ES2640396T3|2017-11-02|
    US10295587B2|2019-05-21|
    US20160003886A1|2016-01-07|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    US20120134058A1|2009-08-14|2012-05-31|Fronius International Gmbh|Method for detecting arcs in photovoltaic systems and such a photovoltaic system|
    FR2983968A1|2011-12-09|2013-06-14|Commissariat Energie Atomique|METHOD AND DEVICE FOR DETECTING ELECTRIC ARC IN A PHOTOVOLTAIC INSTALLATION|
    US6426632B1|1999-03-29|2002-07-30|George A. Spencer|Method and apparatus for testing an AFCI/GFCI circuit breaker|
    US6191589B1|1999-03-29|2001-02-20|George A. Spencer|Test circuit for an AFCI/GFCI circuit breaker|
    US7391218B2|2005-03-11|2008-06-24|Honeywell International Inc.|Method and apparatus for generalized arc fault detection|FR3046232B1|2015-12-28|2018-02-16|Commissariat A L'energie Atomique Et Aux Energies Alternatives|METHOD FOR DETECTING AN ELECTRIC ARC BY ANALYSIS OF ITS ACOUSTIC SIGNATURE|
    FR3053122B1|2016-06-27|2018-07-27|Commissariat A L'energie Atomique Et Aux Energies Alternatives|DEVICE FOR DETECTING AN ELECTRIC ARC FROM ITS ACOUSTIC SIGNATURE|
    CN110456234B|2018-05-07|2020-11-10|珠海格力电器股份有限公司|Fault arc detection method, device and system|
    US10985695B2|2018-08-28|2021-04-20|Analog Devices International Unlimited Company|DC arc detection and photovoltaic plant profiling system|
    法律状态:
    2015-06-24| PLFP| Fee payment|Year of fee payment: 2 |
    2016-01-08| PLSC| Search report ready|Effective date: 20160108 |
    2016-07-01| PLFP| Fee payment|Year of fee payment: 3 |
    2017-07-06| PLFP| Fee payment|Year of fee payment: 4 |
    优先权:
    申请号 | 申请日 | 专利标题
    FR1456477A|FR3023377B1|2014-07-04|2014-07-04|METHOD AND DEVICE FOR DETECTING ELECTRIC ARCS FOR ELECTRICAL INSTALLATIONS WITH A CONTINUOUS CURRENT SOURCE|FR1456477A| FR3023377B1|2014-07-04|2014-07-04|METHOD AND DEVICE FOR DETECTING ELECTRIC ARCS FOR ELECTRICAL INSTALLATIONS WITH A CONTINUOUS CURRENT SOURCE|
    US14/789,183| US10295587B2|2014-07-04|2015-07-01|Method and device for detecting arcing in electrical installation for DC power|
    EP15175200.3A| EP2966461B1|2014-07-04|2015-07-03|Method and device for detecting electrical arcs for electrical systems with a dc source|
    ES15175200.3T| ES2640396T3|2014-07-04|2015-07-03|Procedure and device for detecting electric arcs for electrical installations with direct current source|
    [返回顶部]