GIS equipment partial discharge mode identification method and system, terminal equipment and storage medium

Abstract

The invention discloses a partial discharge mode identification method and system for GIS equipment, terminal equipment and a storage medium, and the method comprises the steps: constructing a multi-dimensional partial discharge feature map according to the discharge data, time domain signals and frequency domain signals of to-be-identified GIS equipment; comprehensive data support is provided for the subsequent partial discharge mode recognition operation of the GIS equipment to be recognized by the partial discharge mode recognition model, so that the partial discharge mode recognition model can comprehensively and comprehensively perform partial discharge mode recognition on the GIS equipment to be recognized, and the accuracy of a partial discharge mode recognition result is improved; the problem that in the prior art, due to the fact that a feature distribution diagram is generated through a single signal, the feature distribution diagram can only reflect partial features of partial discharge, and then the accuracy of a partial discharge mode recognition result obtained by conducting partial discharge mode recognition on the feature distribution diagram through a neural network model is low is solved.

Description

Technical Field

[0001] The present invention relates to the field of local discharge pattern recognition, and in particular to a local discharge pattern recognition method, system, terminal equipment and storage medium for GIS equipment. Background Art

[0002] With the rapid development of grid interconnection, the research and application of new energy power generation technology, low-loss long-distance reliable transmission equipment and grid intelligence have become the focus of major power countries in the world today. Traditional high-voltage AC transmission technology is difficult to adapt to the current needs of modern large-capacity grid interconnection due to its shortcomings such as large line loss, easily affected stability and high line cost, which has greatly restricted the construction and development of the grid. Therefore, high-voltage direct current transmission has been developed. In the actual high-voltage long-distance transmission process, in order to improve the flexibility of high-voltage direct current transmission line corridor selection and meet the reliability of power transmission in harsh environments, related research on high-voltage direct current GIS equipment is also emerging.

[0003] Partial discharge refers to discharge that occurs between electrodes but does not penetrate the electrodes. It is caused by weaknesses in the insulation of the equipment or defects caused during the production process, and repeated breakdown and extinguishing occurs under high electric field strength. Partial discharge is one of the main causes of failure of the main insulation of electrical equipment and an important macroscopic representation of insulation degradation.

[0004] At present, the technology for identifying the partial discharge mode of GIS equipment usually uses the time domain signal or frequency domain signal of partial discharge to generate a characteristic distribution map, and then uses a neural network model to perform partial discharge mode recognition on the characteristic distribution map. However, the characteristic distribution map generated by this method can only reflect some characteristics of partial discharge because it is based on a single signal, which leads to a low accuracy rate of partial discharge mode recognition results obtained by the neural network model for partial discharge mode recognition on the characteristic distribution map. Summary of the invention

[0005] The present invention provides a local discharge pattern recognition method, system, terminal device and storage medium for GIS equipment, which can solve the problem that the existing technology uses a single signal to generate a characteristic distribution map, which results in the characteristic distribution map only reflecting part of the characteristics of the local discharge, and further leads to the problem that the local discharge pattern recognition result obtained by the neural network model for local discharge pattern recognition on the characteristic distribution map has low accuracy.

[0006] In order to solve the above technical problems, an embodiment of the present invention provides a method for identifying a partial discharge pattern of a GIS device, comprising:

[0007] Obtaining discharge data, time domain signals and frequency domain signals of partial discharge of the GIS equipment to be identified; wherein the discharge data includes discharge frequency, discharge time interval and discharge amount;

[0008] Perform time domain feature processing on the time domain signal to obtain the time domain envelope feature of the GIS device to be identified;

[0009] According to the discharge frequency and the time domain envelope characteristics, several significant characteristic parameters are determined, and according to the several significant characteristic parameters, a scatter distribution characteristic diagram of the significant characteristic parameters is constructed;

[0010] According to the discharge frequency, discharge time interval and discharge amount, a two-dimensional discharge distribution characteristic map of the local discharge of the GIS equipment to be identified is constructed, and the discharge density of the local discharge of the GIS equipment to be identified is calculated. According to the two-dimensional discharge distribution characteristic map and the discharge density, a discharge density map is constructed;

[0011] The frequency characteristics of the frequency domain signal are extracted to obtain the discharge frequency domain characteristics of each power generation frequency, and based on the discharge frequency domain characteristics of each power generation frequency, a two-dimensional frequency distribution map of the GIS equipment to be identified is constructed;

[0012] The significant characteristic parameter scatter point distribution characteristic map, discharge density map and two-dimensional frequency distribution map are input into the trained partial discharge recognition model, so that the trained partial discharge recognition model can perform partial discharge pattern recognition according to the significant characteristic parameter scatter point distribution characteristic map, discharge density map and two-dimensional frequency distribution map, and obtain the partial discharge pattern recognition result of the GIS equipment to be identified.

[0013] Furthermore, according to the discharge frequency and the time domain envelope characteristics, several significant characteristic parameters are determined, including:

[0014] According to the discharge frequency and time domain envelope characteristics, the characteristic parameter matrix of the partial discharge signal is constructed, and the information content and information weight of each characteristic parameter are calculated according to the characteristic parameter matrix;

[0015] According to the information content and information weight of each characteristic parameter, several significant characteristic parameters are determined.

[0016] Furthermore, the calculation of the information content and information weight of each characteristic parameter according to the characteristic parameter matrix includes:

[0017] According to the characteristic parameter matrix, the same-class aggregation degree and inter-class dispersion degree of each characteristic parameter are determined, and the information content of each characteristic parameter is calculated according to the same-class aggregation degree and inter-class dispersion degree of each characteristic parameter;

[0018] According to the information amount of each characteristic parameter, the information amount weight of each characteristic parameter is calculated.

[0019] Furthermore, the information volume calculation formula and the information volume weight calculation formula are as follows:

[0020] The information volume calculation formula is:

[0021]

[0022] Among them, J k Represents the amount of information of the kth feature parameter; α k Indicates the similarity aggregation degree of the kth characteristic parameter; β k Represents the inter-class dispersion of the kth feature parameter;

[0023] The information weight calculation formula is:

[0024]

[0025] Among them, W k represents the information weight of the kth feature parameter; K represents the total number of feature parameters; J k Represents the amount of information of the kth feature parameter; Represents the total amount of information of feature parameters.

[0026] Furthermore, the two-dimensional discharge distribution characteristic diagram of the local discharge of the GIS equipment to be identified is constructed according to the discharge frequency, discharge time interval and discharge amount, including:

[0027] According to the discharge frequency, discharge time interval and discharge amount, the maximum discharge amount and the maximum discharge time interval of each discharge frequency are determined, and the discharge amount interval and the discharge time interval interval are constructed;

[0028] Divide the discharge amount interval and the discharge time interval interval into a plurality of sub-intervals, and calculate the mean values ​​of discharge parameters of the previous and next discharge frequencies in each sub-interval; wherein the discharge parameter mean values ​​include the discharge amount mean value and the discharge time interval mean value;

[0029] According to the mean values ​​of the previous and next discharge parameters of each discharge frequency in each subinterval, a two-dimensional discharge distribution characteristic map of the local discharge of the GIS equipment to be identified is constructed.

[0030] Furthermore, the calculation formula for calculating the mean value of the discharge parameters of the previous and next discharge frequencies in each sub-interval is as follows:

[0031] The calculation formula of the previous discharge amount mean value is:

[0032]

[0033] Among them, q aver (i-1, l) represents the mean value of the previous discharge; N l represents the total number of times in the lth interval; It represents the discharge amount of the kth frequency in the previous l subintervals;

[0034] The calculation formula of the latter discharge amount mean value is:

[0035]

[0036] Among them, q aver (i+1,l) represents the average value of the last discharge; N l represents the total number of times in the lth interval; It represents the discharge amount of the kth frequency after I subintervals;

[0037] The calculation formula of the mean value of the previous discharge time interval is:

[0038]

[0039] Among them, Δt aver (i-1,l) represents the mean time interval of the previous discharge; N l represents the total number of times in the lth interval; represents the time interval of the previous discharge of the kth frequency in the lth subinterval;

[0040] The calculation formula of the mean value of the last discharge time interval is:

[0041]

[0042] Among them, Δt aver (i+1,l) represents the mean time interval of the next discharge; N l represents the total number of times in the lth interval; It represents the time interval of a discharge after l subintervals of the kth frequency.

[0043] Furthermore, the partial discharge identification model includes an initial identification module, a BPA calculation module and a final identification module; the initial identification module includes a convolutional neural network unit;

[0044] The trained partial discharge recognition model performs partial discharge pattern recognition according to a significant characteristic parameter scatter point distribution characteristic map, a discharge density map and a two-dimensional frequency distribution map, including:

[0045] The convolutional neural network unit of the initial recognition module performs image feature extraction and discharge pattern recognition on the scatter distribution feature map of significant characteristic parameters, the discharge density map and the two-dimensional frequency distribution map, and obtains the initial recognition results and recognition accuracy of the local discharge pattern in different dimensions; wherein the different dimensions include the time domain dimension, the frequency domain dimension and the time-frequency dimension;

[0046] According to the recognition accuracy of partial discharge patterns in different dimensions, the basic probability value of recognition in each dimension is calculated through the BPA calculation module;

[0047] According to the basic probability value of identification in each dimension, the fusion value of the basic probability value of identification in any two dimensions is calculated through the final identification module, and the initial identification results of the local discharge patterns of the two dimensions with the maximum fusion value are fused to obtain the local discharge pattern recognition result of the GIS equipment to be identified.

[0048] Based on the above method embodiment, the present invention provides a corresponding system embodiment;

[0049] An embodiment of the present invention provides a local discharge pattern recognition system for GIS equipment, comprising: a data acquisition module, a first feature map generation module, a second feature map generation module, a third feature map generation module and a local discharge pattern recognition module;

[0050] The data acquisition module is used to acquire the discharge data, time domain signal and frequency domain signal of the local discharge of the GIS equipment to be identified; wherein the discharge data includes the discharge frequency, the discharge time interval and the discharge amount;

[0051] The first feature map generation module is used to perform time domain feature processing on the time domain signal to obtain the time domain envelope feature of the GIS device to be identified, determine a number of significant feature parameters according to the discharge frequency and the time domain envelope feature, and construct a scatter distribution feature map of the significant feature parameters according to the several significant feature parameters;

[0052] The second characteristic map generation module is used to construct a two-dimensional discharge distribution characteristic map of the local discharge of the GIS equipment to be identified according to the discharge frequency, discharge time interval and discharge amount, and calculate the discharge density of the local discharge of the GIS equipment to be identified, and construct a discharge density map according to the two-dimensional discharge distribution characteristic map and the discharge density;

[0053] The third feature map generation module is used to extract frequency features of the frequency domain signal to obtain the discharge frequency domain features of each power generation frequency, and construct a two-dimensional frequency distribution map of the GIS device to be identified according to the discharge frequency domain features of each power generation frequency;

[0054] The local discharge pattern recognition module is used to input the significant characteristic parameter scatter point distribution characteristic diagram, the discharge density spectrum and the two-dimensional frequency distribution spectrum into the trained local discharge recognition model, so that the trained local discharge recognition model performs local discharge pattern recognition according to the significant characteristic parameter scatter point distribution characteristic diagram, the discharge density spectrum and the two-dimensional frequency distribution spectrum, and obtains the local discharge pattern recognition result of the GIS equipment to be identified.

[0055] Based on the above-mentioned method embodiment, the present invention provides a corresponding terminal device embodiment, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, a local discharge pattern recognition method for GIS equipment as described in the present invention is implemented.

[0056] Based on the above method embodiment, the present invention provides a corresponding computer-readable storage medium embodiment, including: a stored computer program, which controls the device where the computer-readable storage medium is located to execute a local discharge pattern recognition method for GIS equipment as described in the present invention when the computer program is running.

[0057] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0058] The present invention constructs a scatter distribution characteristic diagram of significant characteristic parameters of partial discharge of a DC GIS device to be identified, a discharge density map and a two-dimensional frequency distribution map respectively according to the discharge data, time domain signal and frequency domain signal of partial discharge of a DC GIS device to be identified, and uses the constructed scatter distribution characteristic diagram of significant characteristic parameters, the discharge density map and the two-dimensional frequency distribution map as input data of a partial discharge pattern recognition model, so that the partial discharge pattern recognition model can perform partial discharge pattern recognition on the DC GIS device to be identified, that is, the present invention constructs a multi-dimensional partial discharge characteristic diagram according to the discharge data, the time domain signal and the frequency domain signal, and provides comprehensive data support for the subsequent partial discharge pattern recognition model to perform partial discharge pattern recognition operation on the GIS device to be identified, so that the partial discharge pattern recognition model can comprehensively and comprehensively perform partial discharge pattern recognition on the GIS device to be identified, improve the accuracy of the partial discharge pattern recognition result, and solve the problem that the prior art uses a single signal to generate a characteristic distribution map, which results in the characteristic distribution map only reflecting partial characteristics of the partial discharge, and then results in the low accuracy of the partial discharge pattern recognition result obtained by the neural network model performing partial discharge pattern recognition on the characteristic distribution map. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 : A flow chart of the steps of a method for identifying a partial discharge pattern of a GIS device provided by an embodiment of the present invention;

[0060] Figure 2 : A system structure diagram of a partial discharge pattern recognition system for GIS equipment provided by an embodiment of the present invention;

[0061] Figure 3 : An example diagram of the scatter distribution characteristics of significant characteristic parameters constructed experimentally according to an embodiment of the present invention;

[0062] Figure 4: Example graph of discharge density constructed experimentally for an embodiment of the present invention;

[0063] Figure 5 : An example graph of a two-dimensional frequency distribution constructed experimentally according to an embodiment of the present invention; DETAILED DESCRIPTION

[0064] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0065] In the description of the present invention, it should be understood that the terms "first", "second" and "third" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features.

[0066] Embodiment 1:

[0067] Reference Figure 1 , is a flowchart of a method for identifying a partial discharge pattern of a GIS device provided by an embodiment of the present invention, the method at least comprising the following steps:

[0068] Step S1: obtaining discharge data, time domain signals and frequency domain signals of partial discharge of the GIS equipment to be identified; wherein the discharge data includes discharge frequency, discharge time interval and discharge amount;

[0069] In this embodiment, the discharge data of the local discharge signal of the GIS device to be identified can be collected and acquired by using, but not limited to, the MPD800 local discharge measuring instrument, and the time domain signal and frequency domain signal of the local discharge of the GIS device to be identified can be acquired by using, but not limited to, the UHF antenna.

[0070] Step S2: Perform time domain feature processing on the time domain signal to obtain the time domain envelope feature of the GIS device to be identified;

[0071] In this embodiment, the time domain feature processing is performed on the time domain signal through Hilbert transform-average filtering to obtain the time domain envelope feature of the GIS device to be identified.

[0072] Step S3: determining a number of significant characteristic parameters according to the discharge frequency and the time domain envelope characteristics, and constructing a scatter distribution characteristic diagram of the significant characteristic parameters according to the number of significant characteristic parameters;

[0073] In this embodiment, several significant characteristic parameters are determined according to the discharge frequency and the time domain envelope characteristics, including:

[0074] According to the discharge frequency and time domain envelope characteristics, the characteristic parameter matrix of the partial discharge signal is constructed, and the information content and information weight of each characteristic parameter are calculated according to the characteristic parameter matrix;

[0075] In this embodiment, constructing the characteristic parameter matrix of the partial discharge signal further includes: performing standardization processing on the constructed characteristic parameter matrix of the partial discharge signal;

[0076] According to the information content and information weight of each characteristic parameter, several significant characteristic parameters are determined;

[0077] In this embodiment, the significant feature parameters include but are not limited to mean, variance, skewness, and steepness.

[0078] In this embodiment, the information content and information weight of each feature parameter are calculated according to the feature parameter matrix, including:

[0079] According to the characteristic parameter matrix, the same-class aggregation degree and inter-class dispersion degree of each characteristic parameter are determined, and the information content of each characteristic parameter is calculated according to the same-class aggregation degree and inter-class dispersion degree of each characteristic parameter;

[0080] In this embodiment, the information amount calculation formula is:

[0081]

[0082] Among them, J k Represents the amount of information of the kth feature parameter; α k Indicates the similarity aggregation degree of the kth characteristic parameter; β k represents the inter-class dispersion of the kth feature parameter; J k The smaller the value, the stronger the ability of this characteristic parameter to separate different types of partial discharge signals is;

[0083] According to the information amount of each characteristic parameter, the information amount weight of each characteristic parameter is calculated;

[0084] In this embodiment, the information weight calculation formula is:

[0085]

[0086] Among them, W k represents the information weight of the kth feature parameter; K represents the total number of feature parameters; J k Represents the amount of information of the kth feature parameter; Represents the total amount of information of feature parameters; W k The smaller the value, the stronger the ability of the characteristic parameter to separate different discharge signals and the higher the degree of aggregation of similar discharge signals.

[0087] Step S4: constructing a two-dimensional discharge distribution characteristic map of the local discharge of the GIS equipment to be identified according to the discharge frequency, discharge time interval and discharge amount, and calculating the discharge density of the local discharge of the GIS equipment to be identified, and constructing a discharge density map according to the two-dimensional discharge distribution characteristic map and the discharge density;

[0088] In this embodiment, the two-dimensional discharge distribution characteristic diagram of the local discharge of the GIS equipment to be identified is constructed according to the discharge frequency, discharge time interval and discharge amount, including:

[0089] According to the discharge frequency, discharge time interval and discharge amount, the maximum discharge amount and the maximum discharge time interval of each discharge frequency are determined, and the discharge amount interval and the discharge time interval interval are constructed;

[0090] Divide the discharge amount interval and the discharge time interval interval into a plurality of sub-intervals, and calculate the mean values ​​of discharge parameters of the previous and next discharge frequencies in each sub-interval; wherein the discharge parameter mean values ​​include the discharge amount mean value and the discharge time interval mean value;

[0091] According to the mean values ​​of the previous and next discharge parameters of each discharge frequency in each subinterval, a two-dimensional discharge distribution characteristic map of the local discharge of the GIS equipment to be identified is constructed.

[0092] In this embodiment, the calculation formula of the discharge density is: discharge density=total discharge frequency / discharge time; wherein the total discharge frequency and the discharge time are both the total discharge frequency and the discharge time in each sub-interval.

[0093] In this embodiment, the calculation formula for calculating the mean value of the discharge parameters of the previous and next discharge frequencies in each sub-interval is as follows:

[0094] The calculation formula of the previous discharge amount mean value is:

[0095]

[0096] Among them, q aver (i-1, l) represents the mean value of the previous discharge; N l represents the total number of times in the lth interval; It represents the discharge amount of the kth frequency in the previous l subintervals;

[0097] The calculation formula of the latter discharge amount mean value is:

[0098]

[0099] Among them, q aver (i+1, l) represents the average value of the last discharge; N l represents the total number of times in the lth interval; It represents the discharge amount of the kth frequency after I subintervals;

[0100] The calculation formula of the mean value of the previous discharge time interval is:

[0101]

[0102] Among them, Δt aver (i-1, l) represents the mean time interval of the previous discharge; N l represents the total number of times in the lth interval; represents the time interval of the previous discharge of the kth frequency in the lth subinterval;

[0103] The calculation formula of the mean value of the last discharge time interval is:

[0104]

[0105] Among them, Δt aver (i+1, l) represents the mean time interval of the next discharge; N l represents the total number of times in the lth interval; It represents the time interval of a discharge after l subintervals of the kth frequency.

[0106] Step S5: extracting frequency characteristics of the frequency domain signal to obtain the discharge frequency domain characteristics of each power generation frequency, and constructing a two-dimensional frequency distribution map of the GIS device to be identified according to the discharge frequency domain characteristics of each power generation frequency;

[0107] In this embodiment, the frequency characteristics of the frequency domain signal can be extracted by using Fourier transform, but is not limited to, to obtain the discharge frequency domain characteristics of each power generation frequency.

[0108] In this embodiment, the two-dimensional frequency distribution map of the GIS device to be identified is constructed according to the discharge frequency domain characteristics of each power generation frequency, including:

[0109] According to the discharge frequency domain characteristics of each power generation frequency, the interval range of the frequency distribution spectrum is determined, and according to the discharge frequency domain characteristics and the interval range, a three-dimensional frequency distribution spectrum is constructed;

[0110] The three-dimensional frequency distribution map is subjected to dimensionality reduction processing to obtain a two-dimensional frequency distribution map of the GIS device to be identified.

[0111] Step S6: inputting the significant characteristic parameter scatter point distribution characteristic diagram, the discharge density map and the two-dimensional frequency distribution map into the trained partial discharge recognition model, so that the trained partial discharge recognition model can perform partial discharge pattern recognition according to the significant characteristic parameter scatter point distribution characteristic diagram, the discharge density map and the two-dimensional frequency distribution map, and obtain the partial discharge pattern recognition result of the GIS equipment to be identified.

[0112] In this embodiment, refer to Figure 3 , 4 and 5, respectively, are an example graph of scatter distribution characteristics of significant characteristic parameters, an example graph of discharge density, and an example graph of two-dimensional frequency distribution constructed experimentally according to an embodiment of the present invention; the partial discharge recognition model includes an initial recognition module, a BPA calculation module, and a final recognition module; the initial recognition module includes a convolutional neural network unit;

[0113] The trained partial discharge recognition model performs partial discharge pattern recognition according to a significant characteristic parameter scatter point distribution characteristic map, a discharge density map and a two-dimensional frequency distribution map, including:

[0114] The convolutional neural network unit of the initial recognition module performs image feature extraction and discharge pattern recognition on the scatter distribution feature map of significant characteristic parameters, the discharge density map and the two-dimensional frequency distribution map, and obtains the initial recognition results and recognition accuracy of the local discharge pattern in different dimensions; wherein the different dimensions include the time domain dimension, the frequency domain dimension and the time-frequency dimension;

[0115] According to the recognition accuracy of partial discharge patterns in different dimensions, the basic probability value of recognition in each dimension is calculated through the BPA calculation module;

[0116] According to the basic probability value of identification in each dimension, the fusion value of the basic probability value of identification in any two dimensions is calculated through the final identification module, and the initial identification results of the local discharge patterns of the two dimensions with the maximum fusion value are fused to obtain the local discharge pattern recognition result of the GIS equipment to be identified.

[0117] In this embodiment, the BPA calculation is specifically as follows:

[0118] Assigning an initial confidence level to each hypothesis: For each hypothesis Hi, an initial confidence level can be assigned to it based on prior knowledge or experience, usually represented by a Bel function, denoted as BelHi;

[0119] For each hypothesis Hi, calculate its uncertainty measures Plausibility and Belief:

[0120] PlHi=1-BelθHi;

[0121] Among them, PlHi represents the minimum evidence that can be provided to support the hypothesis Hi is true;

[0122]

[0123] Among them, P(Ek) represents the probability of evidence Ek occurring; BelHi|Ej represents the trust in hypothesis Hi given evidence Ej;

[0124] Update the credibility of each hypothesis based on the Plausibility and Belief values:

[0125] For each Hi∈Θ: 1. Pl_newHiHi=PlHi / ΣjPlHj, update the Plausibility value; 2. Bel_newHi=BelHi / ΣjBelHj, update the Belief value; 3. Iterate through the above steps multiple times until convergence;

[0126] Finally, we get BPA: Finally, we get the final trust distribution BPA of each hypothesis Hi = {Bel_finalΦ, Bel_finalΨ, ...}; where Φ, Ψ, etc. represent specific events or hypotheses.

[0127] In this embodiment, the fusion BPA calculation is specifically as follows:

[0128] For each hypothesis Hi∈Θ: 1. Calculate the joint trust Wherein Bel_1, Bel_2 and Bel_3 represent the trust of three sources respectively; 2. Calculate the normalization factor Z = 1 / 1-∑iPlcomb(Hi); 3. The final joint trust is Bel_finalHi = Z*Bel_combHi.

[0129] In this embodiment, the model training of the partial discharge identification model includes:

[0130] Obtaining historical discharge data, historical time domain signals, and historical frequency domain signals of partial discharge of GIS equipment with defect labels; wherein the historical discharge data includes historical discharge frequency, historical discharge time interval, and historical discharge amount;

[0131] Perform time domain feature processing on historical time domain signals to obtain historical time domain envelope features of GIS equipment with defect labels;

[0132] According to the historical discharge frequency and historical time domain envelope characteristics, several historical significant characteristic parameters of GIS equipment with defect labels are determined, and according to several historical significant characteristic parameters, a scatter distribution characteristic map of historical significant characteristic parameters is constructed;

[0133] According to the historical discharge frequency, historical discharge time interval and historical discharge amount, a historical two-dimensional discharge distribution characteristic map of partial discharge of GIS equipment with defect labels is constructed, and the historical discharge density of partial discharge of GIS equipment with defect labels is calculated. According to the historical two-dimensional discharge distribution characteristic map and historical discharge density, a historical discharge density map is constructed.

[0134] The frequency characteristics of the historical frequency domain signals are extracted to obtain the historical discharge frequency domain characteristics of each power generation frequency. Based on the historical discharge frequency domain characteristics of each power generation frequency, a historical two-dimensional frequency distribution map of GIS equipment with defect labels is constructed.

[0135] Inputting a historical significant characteristic parameter scatter point distribution characteristic graph, a historical discharge density graph and a historical two-dimensional frequency distribution graph into a partial discharge recognition model to be trained, so that the partial discharge recognition model to be trained performs partial discharge pattern recognition according to the historical significant characteristic parameter scatter point distribution characteristic graph, the historical discharge density graph and the historical two-dimensional frequency distribution graph, and obtains a partial discharge pattern recognition result; wherein the partial discharge pattern recognition result is a defect result of the GIS equipment; the defect result of the GIS equipment includes but is not limited to a needle tip defect, a surface contamination defect, a metal particle defect and an air gap defect;

[0136] The partial discharge pattern recognition results are compared with the actual defect labels to calculate the loss value, and the parameters of the partial discharge recognition model are optimized according to the loss value until the loss value converges to obtain a trained partial discharge recognition model.

[0137] In this embodiment, the loss function for calculating the loss value includes but is not limited to the mean square error function and the mean absolute percentage error function; the method for optimizing the parameters of the future carbon emission prediction model to be trained includes but is not limited to the gradient descent method and the Adam optimizer.

[0138] Embodiment 2:

[0139] Reference Figure 2 , is a system structure diagram of a local discharge pattern recognition system for GIS equipment provided by an embodiment of the present invention, the system at least includes: a data acquisition module, a first feature map generation module, a second feature map generation module, a third feature map generation module and a local discharge pattern recognition module;

[0140] The data acquisition module is used to acquire the discharge data, time domain signal and frequency domain signal of the local discharge of the GIS equipment to be identified; wherein the discharge data includes the discharge frequency, the discharge time interval and the discharge amount;

[0141] The first feature map generation module is used to perform time domain feature processing on the time domain signal to obtain the time domain envelope feature of the GIS device to be identified, determine a number of significant feature parameters according to the discharge frequency and the time domain envelope feature, and construct a scatter distribution feature map of the significant feature parameters according to the several significant feature parameters;

[0142] The second characteristic map generation module is used to construct a two-dimensional discharge distribution characteristic map of the local discharge of the GIS equipment to be identified according to the discharge frequency, discharge time interval and discharge amount, and calculate the discharge density of the local discharge of the GIS equipment to be identified, and construct a discharge density map according to the two-dimensional discharge distribution characteristic map and the discharge density;

[0143] The third feature map generation module is used to extract frequency features of the frequency domain signal to obtain the discharge frequency domain features of each power generation frequency, and construct a two-dimensional frequency distribution map of the GIS device to be identified according to the discharge frequency domain features of each power generation frequency;

[0144] The local discharge pattern recognition module is used to input the significant characteristic parameter scatter point distribution characteristic diagram, the discharge density spectrum and the two-dimensional frequency distribution spectrum into the trained local discharge recognition model, so that the trained local discharge recognition model performs local discharge pattern recognition according to the significant characteristic parameter scatter point distribution characteristic diagram, the discharge density spectrum and the two-dimensional frequency distribution spectrum, and obtains the local discharge pattern recognition result of the GIS equipment to be identified.

[0145] Based on the above method embodiment, another embodiment is provided;

[0146] Another embodiment of the present invention provides a local discharge pattern recognition terminal device for GIS equipment, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements a local discharge pattern recognition method for GIS equipment described in any one of the above method embodiments of the present invention.

[0147] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program in the local discharge pattern recognition terminal device of the GIS device.

[0148] The local discharge pattern recognition terminal device of the GIS device can be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The local discharge pattern recognition terminal device of the GIS device can include, but is not limited to, a processor and a memory. Those skilled in the art can understand that, for example, the local discharge pattern recognition terminal device of the GIS device can also include an input and output device, a network access device, a bus, etc.

[0149] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the local discharge pattern recognition terminal device of the GIS device, and uses various interfaces and lines to connect various parts of the local discharge pattern recognition terminal device of the entire GIS device.

[0150] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the local discharge pattern recognition terminal device of the GIS device by running or executing the computer program and / or module stored in the memory and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0151] Based on the above method embodiment, another embodiment is provided;

[0152] A storage medium provided in another embodiment of the present invention includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute a local discharge pattern recognition method for GIS equipment described in any one of the above method embodiments of the present invention.

[0153] Wherein, the above-mentioned storage medium is a computer-readable storage medium. The module / unit integrated with the local discharge pattern recognition system / terminal equipment of the GIS equipment can be stored in a computer-readable storage medium if it is implemented in the form of a software functional unit and sold or used as an independent product. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned method embodiments can be implemented. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium.

[0154] It should be noted that the above-mentioned terminal device may include, but is not limited to, a processor and a memory. Those skilled in the art will understand that the above-mentioned terminal device is merely an example and does not constitute a limitation on the terminal device. It may include more or fewer components, or a combination of certain components, or different components.

[0155] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for identifying partial discharge patterns of GIS equipment, characterized in that: include: Obtaining discharge data, time domain signals and frequency domain signals of partial discharge of the GIS equipment to be identified; wherein the discharge data includes discharge frequency, discharge time interval and discharge amount; Perform time domain feature processing on the time domain signal to obtain the time domain envelope feature of the GIS device to be identified; According to the discharge frequency and the time domain envelope characteristics, several significant characteristic parameters are determined, and according to the several significant characteristic parameters, a scatter distribution characteristic diagram of the significant characteristic parameters is constructed; According to the discharge frequency, discharge time interval and discharge amount, a two-dimensional discharge distribution characteristic map of the local discharge of the GIS equipment to be identified is constructed, and the discharge density of the local discharge of the GIS equipment to be identified is calculated. According to the two-dimensional discharge distribution characteristic map and the discharge density, a discharge density map is constructed; The frequency characteristics of the frequency domain signal are extracted to obtain the discharge frequency domain characteristics of each power generation frequency, and based on the discharge frequency domain characteristics of each power generation frequency, a two-dimensional frequency distribution map of the GIS equipment to be identified is constructed; The significant characteristic parameter scatter point distribution characteristic map, discharge density map and two-dimensional frequency distribution map are input into the trained partial discharge recognition model, so that the trained partial discharge recognition model can perform partial discharge pattern recognition according to the significant characteristic parameter scatter point distribution characteristic map, discharge density map and two-dimensional frequency distribution map, and obtain the partial discharge pattern recognition result of the GIS equipment to be identified.

2. The method for identifying partial discharge patterns of GIS equipment according to claim 1, characterized in that: According to the discharge frequency and the time domain envelope characteristics, several significant characteristic parameters are determined, including: According to the discharge frequency and time domain envelope characteristics, the characteristic parameter matrix of the partial discharge signal is constructed, and the information content and information weight of each characteristic parameter are calculated according to the characteristic parameter matrix; According to the information content and information weight of each characteristic parameter, several significant characteristic parameters are determined.

3. The method for identifying partial discharge patterns of GIS equipment according to claim 2, characterized in that: The step of calculating the information content and information weight of each feature parameter according to the feature parameter matrix includes: According to the characteristic parameter matrix, the same-class aggregation degree and inter-class dispersion degree of each characteristic parameter are determined, and the information content of each characteristic parameter is calculated according to the same-class aggregation degree and inter-class dispersion degree of each characteristic parameter; According to the information amount of each characteristic parameter, the information amount weight of each characteristic parameter is calculated.

4. The method for identifying partial discharge patterns of GIS equipment according to claim 3, characterized in that: The information volume calculation formula and information volume weight calculation formula are as follows: The information volume calculation formula is: Among them, J k Represents the amount of information of the kth feature parameter; α k Indicates the similarity aggregation degree of the kth characteristic parameter; β k Represents the inter-class dispersion of the kth feature parameter; The information weight calculation formula is: Among them, W k represents the information weight of the kth feature parameter; K represents the total number of feature parameters; J k Represents the amount of information of the kth feature parameter; Represents the total amount of information of feature parameters.

5. A method for identifying partial discharge patterns of GIS equipment according to claim 4, characterized in that: The method of constructing a two-dimensional discharge distribution characteristic diagram of the local discharge of the GIS equipment to be identified according to the discharge frequency, discharge time interval and discharge amount includes: According to the discharge frequency, discharge time interval and discharge amount, the maximum discharge amount and the maximum discharge time interval of each discharge frequency are determined, and the discharge amount interval and the discharge time interval interval are constructed; Divide the discharge amount interval and the discharge time interval interval into a plurality of sub-intervals, and calculate the mean values ​​of discharge parameters of the previous and next discharge frequencies in each sub-interval; wherein the discharge parameter mean values ​​include the discharge amount mean value and the discharge time interval mean value; According to the mean values ​​of the previous and next discharge parameters of each discharge frequency in each subinterval, a two-dimensional discharge distribution characteristic map of the local discharge of the GIS equipment to be identified is constructed.

6. A method for identifying partial discharge patterns of GIS equipment according to claim 5, characterized in that: The calculation formula for calculating the mean value of the discharge parameters of the previous and next discharge frequencies in each sub-interval is as follows: The calculation formula of the previous discharge amount mean value is: Among them, q aver (i-1, l) represents the mean value of the previous discharge; N l represents the total number of times in the lth interval; It represents the discharge amount of the kth frequency in the previous l subintervals; The calculation formula of the latter discharge amount mean value is: Among them, q aver (i+1, l) represents the average value of the last discharge; N l represents the total number of times in the lth interval; It represents the discharge amount of the kth frequency after l subintervals; The calculation formula of the mean value of the previous discharge time interval is: Among them, Δt aver (i-1, l) represents the mean time interval of the previous discharge; N l represents the total number of times in the lth interval; represents the time interval of the previous discharge of the kth frequency in the lth subinterval; The calculation formula of the mean value of the last discharge time interval is: Among them, Δt aver (i+1, l) represents the mean time interval of the next discharge; N l represents the total number of times in the lth interval; It represents the time interval of a discharge after l subintervals of the kth frequency.

7. A method for identifying partial discharge patterns of GIS equipment according to claim 6, characterized in that: The partial discharge identification model includes an initial identification module, a BPA calculation module and a final identification module; the initial identification module includes a convolutional neural network unit; The trained partial discharge recognition model performs partial discharge pattern recognition according to a significant characteristic parameter scatter point distribution characteristic map, a discharge density map and a two-dimensional frequency distribution map, including: The convolutional neural network unit of the initial recognition module performs image feature extraction and discharge pattern recognition on the scatter distribution feature map of significant characteristic parameters, the discharge density map and the two-dimensional frequency distribution map, and obtains the initial recognition results and recognition accuracy of the local discharge pattern in different dimensions; wherein the different dimensions include the time domain dimension, the frequency domain dimension and the time-frequency dimension; According to the recognition accuracy of partial discharge patterns in different dimensions, the basic probability value of recognition in each dimension is calculated through the BPA calculation module; According to the basic probability value of identification in each dimension, the fusion value of the basic probability value of identification in any two dimensions is calculated through the final identification module, and the initial identification results of the local discharge patterns of the two dimensions with the maximum fusion value are fused to obtain the local discharge pattern recognition result of the GIS equipment to be identified.

8. A partial discharge pattern recognition system for GIS equipment, characterized in that: include: A data acquisition module, a first characteristic map generation module, a second characteristic map generation module, a third characteristic map generation module and a partial discharge pattern recognition module; The data acquisition module is used to acquire the discharge data, time domain signal and frequency domain signal of the local discharge of the GIS equipment to be identified; wherein the discharge data includes the discharge frequency, the discharge time interval and the discharge amount; The first feature map generation module is used to perform time domain feature processing on the time domain signal to obtain the time domain envelope feature of the GIS device to be identified, determine a number of significant feature parameters according to the discharge frequency and the time domain envelope feature, and construct a scatter distribution feature map of the significant feature parameters according to the several significant feature parameters; The second characteristic map generation module is used to construct a two-dimensional discharge distribution characteristic map of the local discharge of the GIS equipment to be identified according to the discharge frequency, discharge time interval and discharge amount, and calculate the discharge density of the local discharge of the GIS equipment to be identified, and construct a discharge density map according to the two-dimensional discharge distribution characteristic map and the discharge density; The third feature map generation module is used to extract frequency features of the frequency domain signal to obtain the discharge frequency domain features of each power generation frequency, and construct a two-dimensional frequency distribution map of the GIS device to be identified according to the discharge frequency domain features of each power generation frequency; The local discharge pattern recognition module is used to input the significant characteristic parameter scatter point distribution characteristic diagram, the discharge density spectrum and the two-dimensional frequency distribution spectrum into the trained local discharge recognition model, so that the trained local discharge recognition model performs local discharge pattern recognition according to the significant characteristic parameter scatter point distribution characteristic diagram, the discharge density spectrum and the two-dimensional frequency distribution spectrum, and obtains the local discharge pattern recognition result of the GIS equipment to be identified.

9. A local discharge pattern recognition terminal device for GIS equipment, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, it implements a local discharge pattern recognition method for GIS equipment as described in any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute a local discharge pattern recognition method for GIS equipment as described in any one of claims 1 to 7.

Images