Partial discharge hybrid detection method, device, terminal and detection system

Abstract

The invention provides a partial discharge hybrid detection method and device, a terminal and a detection system, and relates to the technical field of partial discharge detection. The method comprises the following steps: collecting partial discharge signals of a to-be-measured local position of a target transformer, wherein the partial discharge signals comprise ultrasonic signals and voltage-to-ground signals; calculating the ultrasonic signal and the voltage-to-ground signal by using a chaotic synchronous detection method to obtain a feature matrix; and inputting the feature matrix into a fault type detection model, and outputting a fault type serial number of the to-be-detected local position of the target transformer, the fault type detection model being constructed based on an extended neural network. The accuracy of the characteristic matrix of the partial discharge signal is improved by using the chaotic synchronous detection method, and the accuracy of partial discharge detection is improved by using the extended neural network.

Description

Technical Field

[0001] The present application relates to the technical field of partial discharge detection, and in particular to a partial discharge hybrid detection method, device, terminal and detection system. Background Art

[0002] Power equipment plays a key role in the power system, and it is crucial to ensure its safe and stable operation. Partial discharge is the main sign and manifestation of transformer insulation degradation. Effective detection of partial discharge is of great practical significance for accurately evaluating the insulation condition of the transformer.

[0003] The existing technical solutions for detecting partial discharge include a power cable partial discharge detection method using the LSTM algorithm, a transformer fault type identification method based on variational mode decomposition (VMD) and one-dimensional convolutional neural network (1D_CNN), and a low-voltage cable partial discharge detection method based on frequency domain reflectometry. The common disadvantages of these three solutions are: 1. Most of the test methods require multiple sensors and signal parameters, which increases the cost and complexity of the system; 2. All detection algorithms cannot accurately extract the features of complex signals. Therefore, a method is urgently needed to improve the accuracy of partial discharge detection. Summary of the invention

[0004] The present application provides a partial discharge hybrid detection method, device, terminal and detection system to solve the problem in the prior art that inaccurate feature signal extraction of partial discharge detection data leads to inaccurate fault type judgment.

[0005] In a first aspect, the present application provides a hybrid partial discharge detection method, comprising:

[0006] Collecting a partial discharge signal at a local position to be measured of a target transformer, wherein the partial discharge signal includes an ultrasonic signal and a voltage-to-ground signal;

[0007] The ultrasonic signal and the voltage signal to ground are calculated by using a chaotic synchronization detection method to obtain a characteristic matrix;

[0008] The characteristic matrix is ​​input into a fault type detection model, and the fault type serial number of the local position to be tested of the target transformer is output. The fault type detection model is constructed based on an extended neural network.

[0009] In a second aspect, the present application provides a hybrid partial discharge detection device, comprising:

[0010] A signal acquisition module, used for acquiring a partial discharge signal at a local position to be measured of a target transformer, wherein the partial discharge signal includes an ultrasonic signal and a voltage signal to ground;

[0011] A matrix calculation module, used to calculate the ultrasonic signal and the voltage signal to the ground by using a chaotic synchronization detection method to obtain a characteristic matrix;

[0012] The type determination module is used to input the characteristic matrix into a fault type detection model and output the fault type serial number of the local position to be tested of the target transformer. The fault type detection model is constructed based on an extended neural network.

[0013] In a third aspect, the present application provides a terminal comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the method described in the first aspect or any possible implementation of the first aspect are implemented.

[0014] In a fourth aspect, the present application provides a detection system, comprising a partial discharge detection circuit and a terminal as described in the third aspect, wherein the partial discharge detection circuit comprises a high-pass filter circuit and a common-phase amplifier circuit.

[0015] The present application provides a partial discharge hybrid detection method, device, terminal and detection system, which collects partial discharge signals of the local position to be tested of the target transformer, the partial discharge signals include ultrasonic signals and voltage signals to the ground; uses a chaotic synchronous detection method to calculate the ultrasonic signals and the voltage signals to the ground to obtain a characteristic matrix; inputs the characteristic matrix into a fault type detection model, outputs the fault type sequence number of the local position to be tested of the target transformer, and the fault type detection model is constructed based on an extended neural network. The present application uses a chaotic synchronous detection method to improve the accuracy of the characteristic matrix of the partial discharge signal, and uses an extended neural network to jointly improve the accuracy of partial discharge detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0017] Figure 1 is a flow chart for implementing the hybrid partial discharge detection method provided in the embodiment of the present application;

[0018] Figure 2 is a partial discharge test flow chart provided in an embodiment of the present application;

[0019] Figure 3 is a circuit diagram of a high-pass filter circuit provided in an embodiment of the present application;

[0020] Figure 4 is a circuit diagram of a common-mode amplifier circuit provided in an embodiment of the present application;

[0021] Figure 5 It is a schematic diagram of the principle results of the extended neural network provided in the embodiment of the present application;

[0022] Figure 6 is a schematic diagram of the suggested extension distance provided in the embodiment of the present application;

[0023] Figure 7 is a structural schematic diagram of a hybrid partial discharge detection device provided in an embodiment of the present application;

[0024] Figure 8 is a schematic diagram of a terminal provided in an embodiment of the present application;

[0025] Fig. 9 It is a schematic diagram of the detection system provided in the embodiment of the present application. DETAILED DESCRIPTION

[0026] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0027] In order to make the purpose, technical solutions and advantages of the present application clearer, specific embodiments will be described below in conjunction with the accompanying drawings.

[0028] Since partial discharge is the main sign and manifestation of transformer insulation degradation, effective detection of partial discharge is of great practical significance for accurately evaluating the insulation condition of the transformer. Existing partial discharge detection methods include the following:

[0029] Solution 1: In order to solve the problem that the traditional partial discharge detection method ignores the noise signal during the analysis process, resulting in low accuracy of the discharge detection results, this paper uses the power cable partial discharge detection method based on the LSTM algorithm to improve the accuracy of partial discharge signal detection.

[0030] Solution 2: A transformer fault type identification method based on variational mode decomposition (VMD) and a stack of convolutional neural networks (ID_CNN) is proposed. First, the application principle of VMD is described, and then a set of identification methods is proposed in combination with ID_CNN. Finally, the feasibility of this method is tested and compared based on fault samples and test data.

[0031] Solution 3: A method for detecting partial discharge of low-voltage cables based on frequency domain reflection method is proposed. This method uses the principle of frequency domain reflection to determine whether there is partial discharge in the cable by measuring the voltage waveform at both ends of the cable and the weak signal generated by partial discharge. First, a reference cable is connected to one end of the cable to be detected, and a high-frequency pulse signal is applied to the reference cable. Then, the voltage waveform at both ends of the cable is recorded using time domain sampling technology, and Fourier transform is performed to obtain its spectrum information. Next, the partial discharge signal is extracted from the spectrum information, and feature analysis is performed, such as amplitude, phase, etc. Finally, based on the feature analysis results, it is determined whether the cable has partial discharge.

[0032] The common disadvantages of the above three solutions are:

[0033] 1. Most test methods require multiple sensors and signal parameters, which increases the cost and complexity of the system.

[0034] 2. The detection algorithm used cannot accurately extract the features of complex signals.

[0035] Based on the above problems, this application proposes a new partial discharge detection method. By adopting the extension neural network (ENN) and the chaos synchronization detection method, a partial discharge detection system is developed. In terms of signal acquisition, a high-frequency current transformer (HFCT), high-pass filtering, in-phase amplification, high-frequency oscilloscope and other hardware circuits are used to measure partial discharge. The extended neural network and the chaos synchronization detection method are combined with the hardware circuit to develop a partial discharge fault diagnosis system. By extracting the local discharge characteristics through the chaos synchronization detection method, and then using the extended neural network to analyze the characteristic values ​​of the local discharge, any subtle changes in the discharge signal can be effectively detected, thereby realizing the detection of partial discharge.

[0036] The specific process of the new partial discharge detection method is as follows Figure 1 As shown, Figure 1 The implementation flow chart of the hybrid partial discharge detection method provided in the embodiment of the present application is described in detail as follows:

[0037] In step 101, a partial discharge signal of a local position to be measured of a target transformer is collected, where the partial discharge signal includes an ultrasonic signal and a voltage signal to ground.

[0038] In the embodiment of the present application, the above-mentioned hardware circuit is used to collect the partial discharge signal of the local position to be measured of the target transformer, wherein the partial discharge signal includes the ultrasonic signal and the voltage signal to the ground at the local position.

[0039] In a possible implementation, collecting the partial discharge signal of the local position to be measured of the target transformer may include:

[0040] The constructed partial discharge detection circuit is used to collect the partial discharge signal of the local position to be tested of the target transformer.

[0041] Optionally, refer to Figure 2 The partial discharge test process includes an autotransformer (i.e., a target transformer), a high-frequency current transformer, a detection circuit (i.e., a built partial discharge detection circuit), an oscilloscope, and a server, so as to realize partial discharge detection. The detection circuit in the process collects the partial discharge signal of the local position to be tested of the target transformer.

[0042] Among them, since there is a lot of noise interference in the actual measurement, in order to avoid distortion of the measured signal, this embodiment uses a high-pass filter to filter the low-frequency noise, because the high-pass filter can selectively pass the signal part higher than the frequency according to the set cut-off frequency, and suppress the signal below the frequency, which helps to eliminate the influence of the low-frequency component on the signal, thereby reducing the distortion of the signal. In addition, in order to extract a clear discharge signal with a better oscillation effect, this embodiment uses a passive filter to make a high-pass filter circuit, so that the appropriate cut-off frequency can be selected according to the signal characteristics to ensure that the filter effectively suppresses the low-frequency part while transmitting the required high-frequency part.

[0043] In addition, the specific structure of the high-pass filter circuit does not necessarily have to include a specific number or model of components, but depends on the designer's requirements for filtering characteristics and the specific application of the circuit. The high-pass filter circuit can adopt different topological structures, such as RC, RL, RCL series, etc. Each structure has its unique advantages and usage scenarios.

[0044] For example, refer to Figure 3 A circuit diagram of a high-pass filter circuit is shown, the high-pass filter circuit includes a first resistor R1, a second resistor R2, a first inductor L1, a second inductor L2, a third inductor L3, a fourth inductor L4, a first capacitor C1, a second capacitor C2 and a third capacitor C3.

[0045] The first end of the first resistor R1 is respectively connected to the first end of the input end of the high-pass filter circuit, the first end of the first inductor L1, and the first end of the first capacitor C1, and the second end of the first resistor R1 is respectively connected to the second end of the input end of the high-pass filter circuit, the second end of the first inductor L1, the second end of the second inductor L2, the second end of the third inductor L3, the second end of the fourth inductor L4, the second end of the second resistor R2, the second end of the output end of the high-pass filter circuit and the ground end.

[0046] The first end of the second resistor R2 is respectively connected to the first end of the output end of the high-pass filter circuit, the first end of the fourth inductor L4 and the second end of the third capacitor C3.

[0047] The second end of the first capacitor C1 is connected to the first end of the second inductor L2 and the first end of the second capacitor C2 respectively.

[0048] The second end of the second capacitor C2 is connected to the first end of the third inductor L3 and the first end of the third capacitor C3 respectively.

[0049] However, the discharge signal output by the high-pass filter circuit is mostly above 10MHz. In order to maintain the amplified high-frequency signal to a certain extent and prevent serious distortion, this embodiment uses a non-inverting current feedback amplifier to form a common-phase amplifier circuit, and uses negative feedback to stabilize the gain of the amplifier. For example, when the gain value is 1, the bandwidth is 350MHz, when the gain value is 10, the bandwidth is 35MHz, and the resistivity range is 1kΩ-100kΩ, which prevents the amplifier from working in a nonlinear region.

[0050] For example, refer to Figure 4 The circuit diagram of the common-mode amplifier circuit shown in FIG. 1 includes an input resistor R T , load resistance R O , feedback resistor R F , Collector resistance R G and power amplifiers.

[0051] The positive input terminal of the power amplifier is connected to the input terminal of the in-phase amplifier circuit and the input resistor R T The first end of the input resistor R T The second end is grounded.

[0052] The negative input terminal of the power amplifier is connected to the collector resistor R G The first end and feedback resistor R F The first end is connected to the collector resistor R G The second end is grounded.

[0053] The output of the power amplifier is connected to the feedback resistor R F The second end of the load resistor R O The first end is connected to the output end of the in-phase amplifier circuit, and the load resistor R O The second end is grounded.

[0054] Among them, in the common-mode amplifier circuit, the size relationship between the resistors is that the resistance value range is 1kΩ-100kΩ.

[0055] The embodiment of the present application can more comprehensively reflect the characteristics of partial discharge by simultaneously collecting the ultrasonic signal and the voltage signal to ground of the local position to be tested of the target transformer. These two signals each have unique characteristics in the partial discharge process, and can complement each other when used in combination, thereby improving the accuracy and reliability of detection.

[0056] In step 102, the ultrasonic signal and the voltage signal to ground are calculated using a chaotic synchronization detection method to obtain a characteristic matrix.

[0057] Among them, the chaos synchronization detection method was proposed by Edward Norton Lorenz, an American mathematician, meteorologist, and pioneer of chaos theory. The theory discusses the unsteady behavior of nonlinear dynamic systems. The chaotic system is divided into two parts, the master system and the slave system. When the initialization of the two parts is different, the running trajectory of the chaotic system is also different. The chaotic dynamic error system collects the dynamic error between the master system and the slave system.

[0058] In the embodiment of the present application, a chaotic synchronization detection method is adopted, the ultrasonic signal collected in step 101 is input into the main system, and the voltage signal to ground collected in step 101 is input into the slave system, and the characteristic matrix is ​​calculated.

[0059] In a possible implementation, the chaotic synchronization detection method is used to calculate the ultrasonic signal and the ground voltage signal to obtain a characteristic matrix, which may include:

[0060] Inputting the ultrasonic signal into the first formula, and calculating the first matrix;

[0061] Input the voltage signal to the ground into the second formula to calculate and obtain the second matrix;

[0062] Subtract the second matrix from the first matrix to calculate the characteristic matrix;

[0063] Among them, the first formula is:

[0064]

[0065] Among them, L master is the first matrix, and are the state vectors of ultrasonic signals, x 1 、x 2 and x 3 All are ultrasonic signals, α, β, and γ are error coefficients;

[0066] The second formula is:

[0067]

[0068] Among them, L slave is the second matrix, and are the state vectors of the voltage signal to ground, y 1 ,y 2 and 3 Both are voltage signals relative to ground.

[0069] Optionally, the ultrasonic signal collected in step 101 is input into the first formula based on the main system In the calculation, the first matrix is ​​obtained, and then the voltage signal to ground collected in step 101 is input into the second formula based on the slave system Finally, the second matrix is ​​subtracted from the first matrix to obtain the characteristic matrix.

[0070] In a possible implementation, calculating the characteristic matrix by subtracting the second matrix from the first matrix may include:

[0071] Input the first matrix and the second matrix into the chaotic dynamics function to obtain the characteristic matrix. The chaotic dynamics function is:

[0072]

[0073] in, is the feature matrix, and are error vectors, e 1 、e 2 and e 3 All are error values. Optionally, the first matrix and the second matrix are input into the chaotic dynamics function to obtain the characteristic matrix

[0074] Among them, α, β, and γ are all error coefficients, for example, they can be set to 10, 28, and 8 / 3 respectively.

[0075] In addition, since the coordinates of the two centers in the dynamic diagram are clearly defined as chaotic eyes and are used as data features of partial discharge, this embodiment uses the error value e 1 and e 2 Generate chaotic dynamic error map.

[0076] The embodiment of the present application utilizes a chaotic synchronization detection method to calculate the collected signal, and can obtain a feature matrix containing rich feature information, which can effectively extract the nonlinear features in the signal, so that the fault type detection model can more accurately identify different types of faults.

[0077] In step 103, the characteristic matrix is ​​input into a fault type detection model, and the fault type serial number of the local position to be tested of the target transformer is output. The fault type detection model is constructed based on an extended neural network.

[0078] Among them, the scalability theory provides a new method for calculating the degree of association for the sorting process. Its concept is to study the scalability of different projects, in which the contradictions between projects are discussed quantitatively and qualitatively. The object-element model is used to represent the information of the project. The quantity and quality of things are represented by object-element transformation, and the impact of these on things is known through the correlation function. The main feature of artificial neural networks is their parallel learning ability, which forms a complex network system by using a large number of artificial neurons to imitate biological neural networks. Their combination provides a new classification and recognition method called Extension Neural Network (ENN).

[0079] In the embodiment of the present application, the characteristic matrix calculated in step 102 is input into the constructed fault type detection model, and the fault type number of the local position to be tested of the target transformer is output. The fault type detection model is constructed based on an extended neural network.

[0080] The embodiment of the present application can realize fast and automatic judgment of fault types by inputting the feature matrix into the fault type detection model constructed based on the extended neural network. The extended neural network has strong learning and generalization capabilities, can process a large amount of data in a short time and give accurate prediction results, thereby significantly improving detection efficiency.

[0081] In a possible implementation, the training process of the fault type detection model may be:

[0082] Obtain the historical feature matrix corresponding to the historical partial discharge signal of the target transformer and the corresponding historical fault type serial number;

[0083] Construct a fault type detection model based on an extended neural network;

[0084] The historical feature matrix corresponding to the historical partial discharge signal of the target transformer is taken as input, and the corresponding historical fault type serial number is taken as output to train the fault type detection model.

[0085] Optionally, historical features corresponding to historical partial discharge signals of the target transformer are obtained, and historical fault type serial numbers corresponding to the historical features are obtained. Then, an extended neural network is used to construct a fault type detection model, and the historical features are used as input, and the corresponding historical fault type serial numbers are used as output to train the fault type detection model.

[0086] Among them, the neural network can be regarded as supervised learning, and its goal is to adjust the weights of the neural network to obtain the best clustering performance or the smallest clustering error. The principle and results of the extended neural network refer to Figure 5 shown.

[0087] Before learning neural networks, we need to define several variables. Let the training set be {X 1 , T 1}, {X 2 , T 2}, ..., {X Q , T Q}, where Q is the total number of training patterns, X is the input vector of the neural network, and T is the corresponding target output. The i-th input is X i ={x i1 , x i2 , …, x in}, where n is the total number of features. To evaluate the learning ability, the error function is explained as follows:

[0088]

[0089] Among them, E t is the error function, t ij is the target jth output of the i-th input pattern, o ij is the actual jth output of the ith input pattern, n c is the learning rate, j is the number of outputs, and i is the training mode.

[0090] The implementation process of the learning algorithm of the extended neural network is as follows:

[0091] Step 1: The connection weights between input data nodes and output data nodes are set according to the range of classical domains that can be directly obtained from experience, or evaluated through the following training data:

[0092]

[0093] in, is the lower limit weight of the jth neuron in the kth layer, is the upper limit weight of the jth neuron in the kth layer, x ij is the value of the input signal, the output or input from the jth neuron in the i-th layer, k = 1, 2, ..., n, i = 1, 2, ..., Q, j = 1, 2, ..., n c .

[0094] Step 2: Read the i-th training pattern and cluster number P:X i ={x i1 , x i2 , …, x in}.

[0095] Step 3: Calculate the input pattern X using the extended distance method i The distance to the kth cluster is as follows:

[0096]

[0097] Among them, ED ik To extend the distance.

[0098] The proposed extended distance is a distance metric such as Figure 6 As shown, the proposed extended distance ED ik can be described as x and a range (w L , w U ), which is different from the traditional Euclidean metric. It can be seen that due to different sensitivities, different ranges of classical domains can reach different distances. This is an important advantage in classification applications. Usually, if the feature covers a large range, the data requires that the sensitivity to distance is vague. However, if the feature covers a small range, the data accuracy requirements and the sensitivity to distance are high.

[0099] Step 4: Find m such that ED im =min{ED ik If m=p, go to step 6, otherwise go to step 5.

[0100] Step 5: Update the weights of the mth and pth clusters as follows:

[0101]

[0102] in, is the new value of the lower limit weight of the th neuron in the mth layer, is the old value of the lower limit weight of the jth neuron in the mth layer, is the new value of the upper limit weight of the th neuron in the mth layer, is the old value of the upper limit weight of the jth neuron in the mth layer, η is the learning rate, which is set to 0.1 in this implementation, j = 1, 2, …

[0103] Step 6: Repeat steps 2 to 5. If all patterns have been classified, the learning cycle ends.

[0104] Step 7: If the total error reaches the preset value or the clustering process has converged, stop; otherwise, return to step 3.

[0105] The present application provides a hybrid partial discharge detection method, which collects partial discharge signals at the local position to be tested of the target transformer, wherein the partial discharge signals include ultrasonic signals and voltage signals to the ground; uses a chaotic synchronous detection method to calculate the ultrasonic signals and the voltage signals to the ground to obtain a characteristic matrix; inputs the characteristic matrix into a fault type detection model, and outputs the fault type sequence number of the local position to be tested of the target transformer, wherein the fault type detection model is constructed based on an extended neural network. The present application improves the accuracy of the characteristic matrix of the partial discharge signal by using a chaotic synchronous detection method, and improves the accuracy of partial discharge detection by using an extended neural network.

[0106] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0107] The following is an embodiment of the device of the present application. For details not described in detail, please refer to the corresponding method embodiment described above.

[0108] Figure 7 The structure diagram of the hybrid detection device for partial discharge provided in the embodiment of the present application is shown. For the convenience of explanation, only the part related to the embodiment of the present application is shown, which is described in detail as follows:

[0109] like Figure 7 As shown, the local discharge hybrid detection device 7 comprises:

[0110] A signal acquisition module 71 is used to acquire a partial discharge signal at a local position to be measured of a target transformer, wherein the partial discharge signal includes an ultrasonic signal and a voltage signal to ground;

[0111] The matrix calculation module 72 is used to calculate the ultrasonic signal and the voltage signal to the ground by using the chaotic synchronization detection method to obtain a characteristic matrix;

[0112] The type determination module 73 is used to input the characteristic matrix into the fault type detection model and output the fault type serial number of the local position to be tested of the target transformer. The fault type detection model is constructed based on the extended neural network.

[0113] The present application provides a hybrid partial discharge detection device, which collects partial discharge signals at the local position to be tested of the target transformer, wherein the partial discharge signals include ultrasonic signals and voltage signals to the ground; uses a chaotic synchronous detection method to calculate the ultrasonic signals and the voltage signals to the ground to obtain a characteristic matrix; inputs the characteristic matrix into a fault type detection model, and outputs the fault type sequence number of the local position to be tested of the target transformer, wherein the fault type detection model is constructed based on an extended neural network. The present application uses a chaotic synchronous detection method to improve the accuracy of the characteristic matrix of the partial discharge signal, and uses an extended neural network to jointly improve the accuracy of partial discharge detection.

[0114] In one possible implementation, the matrix calculation module may be used to:

[0115] Inputting the ultrasonic signal into the first formula, and calculating the first matrix;

[0116] Input the voltage signal to the ground into the second formula to calculate and obtain the second matrix;

[0117] Subtract the second matrix from the first matrix to calculate the characteristic matrix;

[0118] Among them, the first formula is:

[0119]

[0120] Among them, L master is the first matrix, and are the state vectors of ultrasonic signals, x 1 、x 2 and x 3 All are ultrasonic signals, α, β, and γ are error coefficients;

[0121] The second formula is:

[0122]

[0123] Among them, L slave is the second matrix, and are the state vectors of the voltage signal to ground, y 1 ,y 2 and 3 All are voltage signals to ground.

[0124] In one possible implementation, the matrix calculation module may be used to:

[0125] Input the first matrix and the second matrix into the chaotic dynamics function to obtain the characteristic matrix. The chaotic dynamics function is:

[0126]

[0127] in, is the feature matrix, and are error vectors, e 1 、e 2 and e 3 All are error values.

[0128] In a possible implementation, the training process of the fault type detection model may be:

[0129] Obtain the historical feature matrix corresponding to the historical partial discharge signal of the target transformer and the corresponding historical fault type serial number;

[0130] Construct a fault type detection model based on an extended neural network;

[0131] The historical feature matrix corresponding to the historical partial discharge signal of the target transformer is taken as input, and the corresponding historical fault type serial number is taken as output to train the fault type detection model.

[0132] In a possible implementation, the signal acquisition module may be used to:

[0133] The constructed partial discharge detection circuit is used to collect the partial discharge signal of the local position to be tested of the target transformer.

[0134] Figure 8 is a schematic diagram of a terminal provided in an embodiment of the present application. Figure 8 As shown, the terminal 8 of this embodiment includes: a processor 80, a memory 81, and a computer program 82 stored in the memory 81 and executable on the processor 80. When the processor 80 executes the computer program 82, the steps in the above-mentioned partial discharge hybrid detection method embodiments are implemented, for example Figure 1 Alternatively, when the processor 80 executes the computer program 82, the functions of each module / unit in the above-mentioned device embodiments are realized, for example, Figure 7 The functions of each module are shown.

[0135] Exemplarily, the computer program 82 may be divided into one or more modules / units, which are stored in the memory 81 and executed by the processor 80 to complete the present application. 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 82 in the terminal 8. For example, the computer program 82 may be divided into Figure 7 The modules shown.

[0136] The terminal 8 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The terminal 8 may include, but is not limited to, a processor 80 and a memory 81. Those skilled in the art will appreciate that Figure 8 It is only an example of terminal 8 and does not constitute a limitation on terminal 8. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the terminal may also include input and output devices, network access devices, buses, etc.

[0137] The processor 80 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. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0138] The memory 81 may be an internal storage unit of the terminal 8, such as a hard disk or memory of the terminal 8. The memory 81 may also be an external storage device of the terminal 8, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the terminal 8. Further, the memory 81 may also include both an internal storage unit of the terminal 8 and an external storage device. The memory 81 is used to store the computer program and other programs and data required by the terminal. The memory 81 may also be used to temporarily store data that has been output or is to be output.

[0139] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0140] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0141] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0142] In the embodiments provided in the present application, it should be understood that the disclosed devices / terminals and methods can be implemented in other ways. For example, the device / terminal embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0143] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0144] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0145] This application also proposes a detection system, referring to Fig. 9 The detection system includes a partial discharge detection circuit 91 and Figure 8 The terminal shown in FIG. 9 is a schematic diagram of a partial discharge detection circuit 91 .

[0146] In one possible implementation, referring to Figure 3 , the high-pass filter circuit 911 may include a first resistor, a second resistor, a first inductor, a second inductor, a third inductor, a fourth inductor, a first capacitor, a second capacitor and a third capacitor;

[0147] The first end of the first resistor is respectively connected to the first end of the input end of the high-pass filter circuit, the first end of the first inductor, and the first end of the first capacitor, and the second end of the first resistor is respectively connected to the second end of the input end of the high-pass filter circuit, the second end of the first inductor, the second end of the second inductor, the second end of the third inductor, the second end of the fourth inductor, the second end of the second resistor, the second end of the output end of the high-pass filter circuit, and the ground end;

[0148] The first end of the second resistor is respectively connected to the first end of the output end of the high-pass filter circuit, the first end of the fourth inductor and the second end of the third capacitor;

[0149] The second end of the first capacitor is connected to the first end of the second inductor and the first end of the second capacitor respectively;

[0150] The second end of the second capacitor is connected to the first end of the third inductor and the first end of the third capacitor respectively.

[0151] In one possible approach, reference Figure 4 , the in-phase amplifier circuit 912 may include an input resistor, a load resistor, a feedback resistor, a collector resistor and a power amplifier;

[0152] The positive input terminal of the power amplifier is connected to the input terminal of the in-phase amplifier circuit and the first terminal of the input resistor respectively, and the second terminal of the input resistor is grounded;

[0153] The negative input terminal of the power amplifier is connected to the first end of the collector resistor and the first end of the feedback resistor respectively, and the second end of the collector resistor is grounded;

[0154] The output end of the power amplifier is respectively connected to the second end of the feedback resistor, the first end of the load resistor and the output end of the in-phase amplifier circuit, and the second end of the load resistor is grounded.

[0155] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application 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 partial discharge hybrid detection method embodiments can be implemented. Among them, 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 that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practices in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practices, computer-readable media does not include electrical carrier signals and telecommunication signals.

[0156] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A hybrid partial discharge detection method, characterized in that: include: Collecting a partial discharge signal at a local position to be measured of a target transformer, wherein the partial discharge signal includes an ultrasonic signal and a voltage-to-ground signal; The ultrasonic signal and the voltage signal to ground are calculated by using a chaotic synchronization detection method to obtain a characteristic matrix; The characteristic matrix is ​​input into a fault type detection model, and the fault type serial number of the local position to be tested of the target transformer is output. The fault type detection model is constructed based on an extended neural network.

2. The hybrid partial discharge detection method according to claim 1, characterized in that: The chaotic synchronization detection method is used to calculate the ultrasonic signal and the voltage signal to the ground to obtain a characteristic matrix, including: Inputting the ultrasonic signal into a first formula to calculate a first matrix; Inputting the voltage signal to ground into a second formula to calculate a second matrix; Subtracting the second matrix from the first matrix to calculate the characteristic matrix; Among them, the first formula is: Among them, L master is the first matrix, and are all state vectors of ultrasonic signals, x1, x2 and x3 are all said ultrasonic signals, α, β and γ are all error coefficients; The second formula is: Among them, L slave is the second matrix, and All are state vectors of voltage signals to ground, and y1, y2 and y3 are all voltage signals to ground.

3. The hybrid partial discharge detection method according to claim 2, characterized in that: The step of subtracting the second matrix from the first matrix to calculate the characteristic matrix includes: The first matrix and the second matrix are input into a chaotic dynamics function to obtain the characteristic matrix. The chaotic dynamics function is: in, is the feature matrix, and are all error vectors, and e1, e2 and e3 are all error values.

4. The hybrid partial discharge detection method according to claim 1, characterized in that: The training process of the fault type detection model is: Obtain the historical feature matrix corresponding to the historical partial discharge signal of the target transformer and the corresponding historical fault type serial number; Building a fault type detection model based on the extended neural network; The historical feature matrix corresponding to the historical partial discharge signal of the target transformer is used as input, and the corresponding historical fault type serial number is used as output to train the fault type detection model.

5. The hybrid partial discharge detection method according to claim 1, characterized in that: The collecting of the partial discharge signal of the local position to be measured of the target transformer includes: The constructed partial discharge detection circuit is used to collect the partial discharge signal of the local position to be tested of the target transformer.

6. A hybrid partial discharge detection device, characterized in that: include: A signal acquisition module, used for acquiring a partial discharge signal at a local position to be measured of a target transformer, wherein the partial discharge signal includes an ultrasonic signal and a voltage signal to ground; A matrix calculation module, used to calculate the ultrasonic signal and the voltage signal to the ground by using a chaotic synchronization detection method to obtain a characteristic matrix; The type determination module is used to input the characteristic matrix into a fault type detection model and output the fault type serial number of the local position to be tested of the target transformer. The fault type detection model is constructed based on an extended neural network.

7. A terminal comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the partial discharge hybrid detection method as claimed in any one of claims 1 to 5 are implemented.

8. A detection system, characterized in that: It comprises a partial discharge detection circuit and the terminal as claimed in claim 7, wherein the partial discharge detection circuit comprises a high-pass filter circuit and a common-phase amplifier circuit.

9. The detection system according to claim 8, characterized in that: The high-pass filter circuit includes a first resistor, a second resistor, a first inductor, a second inductor, a third inductor, a fourth inductor, a first capacitor, a second capacitor and a third capacitor; The first end of the first resistor is respectively connected to the first end of the input end of the high-pass filter circuit, the first end of the first inductor, and the first end of the first capacitor, and the second end of the first resistor is respectively connected to the second end of the input end of the high-pass filter circuit, the second end of the first inductor, the second end of the second inductor, the second end of the third inductor, the second end of the fourth inductor, the second end of the second resistor, the second end of the output end of the high-pass filter circuit, and the ground end; The first end of the second resistor is respectively connected to the first end of the output end of the high-pass filter circuit, the first end of the fourth inductor and the second end of the third capacitor; The second end of the first capacitor is connected to the first end of the second inductor and the first end of the second capacitor respectively; The second end of the second capacitor is connected to the first end of the third inductor and the first end of the third capacitor respectively.

10. The detection system according to claim 8, characterized in that: The in-phase amplifier circuit comprises an input resistor, a load resistor, a feedback resistor, a collector resistor and a power amplifier; The positive input terminal of the power amplifier is connected to the input terminal of the in-phase amplifier circuit and the first terminal of the input resistor respectively, and the second terminal of the input resistor is grounded; The negative input terminal of the power amplifier is connected to the first end of the collector resistor and the first end of the feedback resistor respectively, and the second end of the collector resistor is grounded; The output end of the power amplifier is respectively connected to the second end of the feedback resistor, the first end of the load resistor and the output end of the in-phase amplifier circuit, and the second end of the load resistor is grounded.

Images