Transmission system fault diagnosis method, device, computer equipment and storage medium

Abstract

The present application discloses a method, device, computer equipment and storage medium for diagnosing power transmission system faults. The method includes: collecting simulation fault data when different types of faults occur in multiple simulations of a high-voltage direct current transmission system simulation model; decomposing the simulation fault data of each simulation to obtain multiple components; constructing a fault feature vector based on multiple components of the same group; training a high-voltage direct current transmission system fault diagnosis model using a sample set constructed by the fault feature vector; obtaining target fault data when a fault occurs in an actually operating high-voltage direct current transmission system; decomposing the target fault data to obtain target components; forming a fault feature vector based on the target components and inputting it into a trained high-voltage direct current transmission system fault diagnosis model to obtain the fault type of the fault that occurs in the actually operating high-voltage direct current transmission system. The present application can accurately distinguish between DC line faults in a high-voltage direct current transmission system and AC system faults on the inverter side.

Description

Technical Field

[0001] The present application relates to the technical field of power system fault diagnosis and identification, and in particular to a method, device, computer equipment and storage medium for diagnosing power transmission system faults. Background Art

[0002] The Yongfu DC transmission system is a provincial DC transmission system recently built in my country. Since its inverter-side power supply is a weak AC system, if the fault type cannot be accurately judged and corresponding measures cannot be taken when various faults occur, it will cause serious harm and great instability to the transmission system.

[0003] However, the electrical signal changes of inverter-side AC system faults and DC transmission line faults are extremely similar, and it is difficult to determine the fault type only by the changing characteristics of the electrical signals after the fault.

[0004] Moreover, at present, the fault diagnosis and identification of high-voltage direct current transmission systems is mostly done by manually extracting fault feature information representing different fault types from a number of fault electrical quantities based on experience, analyzing the fault feature information, and deriving the fault type. This traditional fault type identification method starts with theoretical analysis and manual extraction of fault features of various electrical quantities. The implementation process is difficult, and the fault identification rate and accuracy are low. Summary of the invention

[0005] The main purpose of the present application is to provide a method, device, computer equipment and storage medium for diagnosing power transmission system faults, which can solve the technical problem of low recognition rate and accuracy of manual power transmission system fault judgment in the prior art.

[0006] To achieve the above objectives, the present application provides a first aspect of a method for diagnosing a power transmission system fault, the method comprising:

[0007] Collect simulated fault data when different types of faults occur in multiple simulations of the high-voltage direct current transmission system simulation model, where the simulated fault data includes simulated fault voltage data of the target sampling point, and different types of faults include DC line faults and single-phase grounding faults, two-phase short-circuit grounding faults, and three-phase grounding faults of the inverter-side AC system;

[0008] Decomposing the simulated fault data obtained from each simulation using a target decomposition algorithm to obtain corresponding multiple components;

[0009] According to multiple different components corresponding to the same set of simulated fault data, a corresponding fault feature vector is constructed, and according to multiple groups of different fault feature vectors corresponding to different fault types and corresponding known fault types, a sample set is constructed;

[0010] The pre-trained HVDC system fault diagnosis model is trained by using the sample set to obtain a trained HVDC system fault diagnosis model;

[0011] Acquire target fault data collected at a target detection location when a fault occurs in an actually operating high voltage direct current transmission system;

[0012] Decompose the target fault data using the target decomposition algorithm to obtain the corresponding target components;

[0013] The target fault feature vector is composed according to the target components and input into the trained HVDC system fault diagnosis model to obtain the fault type of the actually operating HVDC system.

[0014] To achieve the above-mentioned object, the second aspect of the present application provides a high-voltage direct current transmission system fault diagnosis device, the device comprising:

[0015] The first data acquisition module is used to collect simulated fault data when different types of faults occur in multiple simulations of the high-voltage direct current transmission system simulation model, wherein the simulated fault data includes simulated fault voltage data of the target sampling point, and the different types of faults include DC line faults and single-phase grounding faults, two-phase short-circuit grounding faults and three-phase grounding faults of the inverter-side AC system;

[0016] A decomposition module, used for decomposing the simulation fault data obtained from each simulation by using a target decomposition algorithm to obtain corresponding multiple components;

[0017] A sample construction module is used to construct a corresponding fault feature vector according to multiple different components corresponding to the same set of simulated fault data, and to construct a sample set according to multiple groups of different fault feature vectors corresponding to different fault types and corresponding known fault types;

[0018] A model training module, used for training a pre-trained high-voltage direct current transmission system fault diagnosis model using a sample set to obtain a trained high-voltage direct current transmission system fault diagnosis model;

[0019] The second data acquisition module is used to acquire target fault data collected at a target detection position when a fault occurs in the actually operating high voltage direct current transmission system;

[0020] The decomposition module is also used to decompose the target fault data using a target decomposition algorithm to obtain corresponding target components;

[0021] The diagnosis module is used to form a target fault feature vector according to target components and input it into a trained high-voltage direct current transmission system fault diagnosis model to obtain the fault type of the fault occurring in the actually operating high-voltage direct current transmission system.

[0022] To achieve the above-mentioned purpose, the third aspect of the present application provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor performs the following steps:

[0023] Collect simulated fault data when different types of faults occur in multiple simulations of the high-voltage direct current transmission system simulation model, where the simulated fault data includes simulated fault voltage data of the target sampling point, and different types of faults include DC line faults and single-phase grounding faults, two-phase short-circuit grounding faults, and three-phase grounding faults of the inverter-side AC system;

[0024] Decomposing the simulated fault data obtained from each simulation using a target decomposition algorithm to obtain corresponding multiple components;

[0025] According to multiple different components corresponding to the same set of simulated fault data, a corresponding fault feature vector is constructed, and according to multiple groups of different fault feature vectors corresponding to different fault types and corresponding known fault types, a sample set is constructed;

[0026] The pre-trained HVDC system fault diagnosis model is trained by using the sample set to obtain a trained HVDC system fault diagnosis model;

[0027] Acquire target fault data collected at a target detection location when a fault occurs in an actually operating high voltage direct current transmission system;

[0028] Decompose the target fault data using the target decomposition algorithm to obtain the corresponding target components;

[0029] The target fault feature vector is composed according to the target components and input into the trained HVDC system fault diagnosis model to obtain the fault type of the actually operating HVDC system.

[0030] To achieve the above-mentioned purpose, the fourth aspect of the present application provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the following steps:

[0031] Collect simulated fault data when different types of faults occur in multiple simulations of the high-voltage direct current transmission system simulation model, where the simulated fault data includes simulated fault voltage data of the target sampling point, and different types of faults include DC line faults and single-phase grounding faults, two-phase short-circuit grounding faults, and three-phase grounding faults of the inverter-side AC system;

[0032] Decomposing the simulated fault data obtained from each simulation using a target decomposition algorithm to obtain corresponding multiple components;

[0033] According to multiple different components corresponding to the same set of simulated fault data, a corresponding fault feature vector is constructed, and according to multiple groups of different fault feature vectors corresponding to different fault types and corresponding known fault types, a sample set is constructed;

[0034] The pre-trained HVDC system fault diagnosis model is trained by using the sample set to obtain a trained HVDC system fault diagnosis model;

[0035] Acquire target fault data collected at a target detection location when a fault occurs in an actually operating high voltage direct current transmission system;

[0036] Decompose the target fault data using the target decomposition algorithm to obtain the corresponding target components;

[0037] The target fault feature vector is composed according to the target components and input into the trained HVDC system fault diagnosis model to obtain the fault type of the actually operating HVDC system.

[0038] The embodiments of the present application have the following beneficial effects:

[0039] This application can effectively solve the technical problems of manually extracting fault features and manually making fault judgments that lead to inaccurate fault judgments in traditional fault identification methods, that is, it can efficiently and accurately identify different fault types such as DC line faults in high-voltage DC transmission systems and single-phase grounding faults, two-phase short-circuit grounding faults and three-phase grounding faults in inverter-side AC systems. It reduces manual participation, fully automates monitoring and diagnosing transmission system faults, is simple and convenient to implement, and has a high fault recognition rate and a low error rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, 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 work.

[0041] in:

[0042] Figure 1 A flow chart of a method for diagnosing a power transmission system fault in an embodiment of the present application;

[0043] Figure 2 A schematic diagram of a high voltage direct current transmission system simulation model in an embodiment of the present application;

[0044] Figure 3 A schematic diagram of different types of fault settings in simulation in an embodiment of the present application;

[0045] Figure 4 This is a structural block diagram of a device for diagnosing power transmission system faults in one embodiment of the present application;

[0046] Figure 5 It is a structural block diagram of a computer device in an embodiment of the present application. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0048] like Figure 1 As shown, in one embodiment, a method for diagnosing a power transmission system fault is provided. This embodiment is illustrated by applying it to a computer device. The method for diagnosing a power transmission system fault specifically includes the following steps:

[0049] Step S100: Collecting simulated fault data when different types of faults occur in multiple simulations of a high-voltage direct current transmission system simulation model, wherein the simulated fault data includes simulated fault voltage data of a target sampling point, and different types of faults include a DC line fault and a single-phase grounding fault, a two-phase short-circuit grounding fault, and a three-phase grounding fault of an inverter-side AC system.

[0050] Specifically, the high-voltage direct current transmission system simulation model is a Yongfu weak-receiving end high-voltage direct current transmission system simulation model built through the PSCAD / EMTDC platform. Among them, PSCAD (full name Power Systems Computer Aided Design) is a widely used electromagnetic transient simulation software in the world, and EMTDC (full name Electromagnetic Transients including DC) is its simulation calculation core. PSCAD provides a graphical operation interface for EMTDC.

[0051] By setting parameters for the HVDC transmission system simulation model, the HVDC transmission system simulation model can simulate the HVDC transmission system in different working states and obtain simulation data. For example, by setting parameters and setting different faults in the simulation, the HVDC transmission system simulation model can simulate the HVDC transmission system under different faults and obtain electrical experimental data reflecting different fault types, i.e., simulated fault data.

[0052] When a fault occurs in the actual transmission system, it is temporarily impossible to diagnose the specific fault type. Therefore, a large amount of simulated fault data corresponding to each fault type can be obtained through the simulation model, which is simple and efficient for training the fault diagnosis model of the high-voltage direct current transmission system.

[0053] Step S200: Decomposing the simulated fault data obtained in each simulation by using a target decomposition algorithm to obtain corresponding multiple components.

[0054] Specifically, the simulated fault data is a complex signal, which contains rich information reflecting the fault type. However, it is difficult to directly analyze the number of simulated faults, so it needs to be decomposed and simplified.

[0055] The simulated fault data can be decomposed using one of the decomposition methods, such as Empirical Mode Decomposition (EMD), Ensemble Empirical Mode Decomposition (EEMD), Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN), Variational Mode Decomposition (VMD), and wavelet decomposition.

[0056] The simulated fault data obtained from each fault simulation needs to be decomposed, so each set of simulated fault data corresponds to multiple different components. The number of components obtained by decomposing simulated fault data of different fault types may not be the same, and the types of components may not be the same.

[0057] Step S300: construct corresponding fault feature vectors according to multiple components corresponding to the same set of simulated fault data, and construct sample sets according to multiple groups of different fault feature vectors corresponding to different fault types and corresponding known fault types.

[0058] Specifically, multiple different components corresponding to the same set of simulated fault data can be used as fault features to construct a fault feature vector; other result values ​​can also be calculated based on multiple different components corresponding to the same set of simulated fault data, and the corresponding multiple result values ​​can be used as fault features to construct a fault feature vector.

[0059] Each fault type will correspond to multiple simulations, each simulation will generate simulated fault data, and each set of simulated fault data corresponds to a fault feature vector. Therefore, a sample set can be constructed based on the fault feature vectors corresponding to all simulated fault data and the known fault types corresponding to the simulated fault data.

[0060] The sample set can be divided into a training set and a test set in proportion. The training set is used to train the HVDC system fault diagnosis model, and the test set is used to test the model function of the trained HVDC system fault diagnosis model.

[0061] Step S400: training a pre-trained HVDC system fault diagnosis model using a sample set to obtain a trained HVDC system fault diagnosis model.

[0062] Specifically, the sample set includes a plurality of training samples, each training sample includes a fault feature vector corresponding to a set of simulated fault data, and a known fault type thereof as a corresponding label.

[0063] Input a training sample to the pre-trained HVDC system fault diagnosis model. The pre-trained HVDC system fault diagnosis model predicts and diagnoses the fault type of the training sample according to the fault feature vector of the training sample to obtain the diagnosed fault type. The loss function is calculated according to the diagnosed fault type and the known fault type. The model gradient is calculated according to the loss function. The model parameters of the pre-trained HVDC system fault diagnosis model are updated according to the model gradient. Input another training sample to the pre-trained HVDC system fault diagnosis model of the previous round of model parameter update. Repeat the above steps until the HVDC system fault diagnosis model converges. The convergence condition is that the loss function is less than the preset value, or the number of training times reaches the preset number.

[0064] Step S500: acquiring target fault data collected at a target detection position when a fault occurs in an actually operating high voltage direct current power transmission system.

[0065] Specifically, the target detection position is a position in the actually operating high voltage direct current transmission system corresponding to the target sampling point in the high voltage direct current transmission system simulation model. A sampling device is provided at the target detection position of the actually operating high voltage direct current transmission system, and the sampling device sends or transmits the collected target fault data to the computer device.

[0066] The purpose of training the HVDC system fault diagnosis model is to diagnose actual transmission system faults. Therefore, after the model training is completed, the target fault data of the actual HVDC system fault is used as the input of the trained HVDC system fault diagnosis model.

[0067] Step S600: Decompose the target fault data using a target decomposition algorithm to obtain corresponding target components.

[0068] Specifically, both simulated fault data and actual target fault data require signal decomposition.

[0069] Step S700: forming a target fault feature vector according to target components and inputting it into a trained HVDC system fault diagnosis model to obtain the fault type of the fault actually occurring in the HVDC system.

[0070] Specifically, after the target fault data is decomposed, multiple target components are obtained. A target fault feature vector is constructed based on the multiple target components as the input of the trained high-voltage direct current transmission system fault diagnosis model. The trained high-voltage direct current transmission system fault diagnosis model analyzes and diagnoses the fault types of faults occurring in the actually operating high-voltage direct current transmission system based on the target fault feature vector.

[0071] This embodiment can effectively solve the technical problems of manually extracting fault features and manually performing fault judgment in the traditional fault identification method, which leads to inaccurate fault judgment. That is, it can efficiently and accurately distinguish different fault types such as DC line faults in high-voltage DC transmission systems and single-phase grounding faults, two-phase short-circuit grounding faults and three-phase grounding faults in inverter-side AC systems. It reduces manual participation, fully automates monitoring and diagnosing transmission system faults, is simple and convenient to implement, and has a high fault recognition rate and a low error rate.

[0072] In one embodiment, step S100 specifically includes:

[0073] Collect simulated fault voltage data of a target sampling point when a DC line fault occurs on a DC transmission simulation line in each simulation of the high voltage DC transmission system simulation model in a first preset number of times, and obtain simulated fault data corresponding to each DC line fault, wherein at least one of a fault distance and a resistance value of a transition resistance in different rounds of simulation in the first preset number of times of simulation is different;

[0074] Collect simulated fault voltage data of a target sampling point when a single-phase grounding fault occurs in the inverter-side AC simulation system of each high-voltage direct current transmission system simulation model in the second preset number of simulations, and obtain simulated fault data corresponding to each single-phase grounding fault of the inverter-side AC system, wherein the resistance values ​​of the transition resistors in different rounds of simulation in the second preset number of simulations are different;

[0075] Collect simulated fault voltage data of a target sampling point when a two-phase short-circuit grounding fault occurs in the inverter-side AC simulation system of each high-voltage direct current transmission system simulation model in the third preset number of simulations, and obtain simulated fault data corresponding to each two-phase short-circuit grounding fault of the inverter-side AC system, wherein the resistance values ​​of the transition resistors in different rounds of simulation in the third preset number of simulations are different;

[0076] Collect simulated fault voltage data of the target sampling point when a three-phase grounding fault occurs in the inverter-side AC simulation system of each high-voltage direct current transmission system simulation model in the fourth preset number of simulations, and obtain a simulated fault signal corresponding to each three-phase grounding fault of the inverter-side AC system, wherein the resistance values ​​of the transition resistances simulated in different rounds of the fourth preset number of simulations are different.

[0077] Specifically, the fault types that can be diagnosed in the present application include DC line faults and single-phase grounding faults, two-phase short-circuit grounding faults and three-phase grounding faults of the inverter-side AC system.

[0078] Each fault type corresponds to multiple simulations, and the first preset number of times, the second preset number of times, the third preset number of times, and the fourth preset number of times may be equal or different, depending on the actual application scenario configuration.

[0079] The high-voltage direct current transmission system simulation model includes a rectifier-side AC simulation system, a DC transmission simulation line, and an inverter-side AC simulation system connected in sequence. For a DC line fault, a DC line fault is set on the DC transmission simulation line of the high-voltage direct current transmission system simulation model by setting simulation parameters. For the inverter-side AC system, a corresponding fault is set on the inverter-side AC simulation system of the high-voltage direct current transmission system simulation model by setting simulation parameters.

[0080] In order to simulate various types of faults in the HVDC transmission system, different parameters will be set for the HVDC transmission system simulation model to achieve the corresponding faults during simulation. In addition, for different simulations of the same fault type, different settings can be made for the fault distance or over-resistance while ensuring the same fault occurs, so as to reflect the different degrees of fault effects of the same fault.

[0081] Because the setting of the transition resistance cannot be too large, the number of simulation data sets obtained by only changing the transition resistance value is small, which is not conducive to neural network training and learning. If you want to obtain more sets of simulation fault data of DC line grounding faults, you need to change the fault distance to obtain more sets of fault data. That is, set a grounding fault on the DC transmission line, and achieve different degrees of DC line faults by changing parameters such as fault distance and transition resistance. For example, the fault distance can be set in a 577km DC transmission line with a change step of 50km. The value range of the transition resistance is 0.01~400Ω, and the change step is 40Ω.

[0082] In the inverter side AC system, single-phase grounding, two-phase short-circuit grounding and three-phase grounding faults are set respectively, and corresponding faults of different degrees can be realized by changing the transition resistance. For example, the value range of the transition resistance is 0.01~400Ω, and the change step is 40Ω.

[0083] Of course, the above-mentioned rules for changing the fault distance and the excessive resistance can be configured according to actual conditions, and this application does not impose any limitation on this.

[0084] In addition, in the second preset number of simulations, the fault distances in different rounds of simulation are the same, and the resistance values ​​of the transition resistance are different; in the third preset number of simulations, the fault distances in different rounds of simulation are the same, and the resistance values ​​of the transition resistance are different; in the fourth preset number of simulations, the fault distances in different rounds of simulation are the same, and the resistance values ​​of the transition resistance are different.

[0085] Alternatively, at least one of the fault distances and the resistance values ​​of the transition resistance in different rounds of simulation in the second preset number of simulations is different; at least one of the fault distances and the resistance values ​​of the transition resistance in different rounds of simulation in the third preset number of simulations is different; at least one of the fault distances and the resistance values ​​of the transition resistance in different rounds of simulation in the fourth preset number of simulations is different.

[0086] In addition, each simulation obtains a set of simulated fault data, which includes the simulated fault voltages of the same target sampling point at multiple different times during the simulation process, and the simulated fault voltages at these times constitute the simulated fault voltage data of the target sampling point. Adjacent sampling moments at different times can be determined according to the sampling frequency, and the number of sampling times can also be configured according to the actual application scenario.

[0087] In one embodiment, the high-voltage direct current transmission system simulation model includes a rectifier-side AC simulation system, a DC transmission simulation line, and an inverter-side AC simulation system connected in sequence;

[0088] The target sampling point is located on the DC transmission simulation line.

[0089] Specifically, Figure 2 Schematic diagram of a high voltage direct current transmission system simulation model in an embodiment of the present application; Figure 2 The rectifier side AC simulation system 1 includes: a rectifier side AC filter 13, a rectifier side converter transformer 11, and a rectifier side converter valve 12. The rectifier side AC filter 13 is used to filter the rectifier side, the rectifier side converter transformer 11 is used to transform the initial AC power flowing in, and the rectifier side converter valve 12 is used to convert the transformed AC power into DC power.

[0090] The DC power obtained by the rectifier-side AC simulation system 1 is input to the inverter-side AC simulation system 3 through a DC transmission simulation line. The DC transmission simulation line includes a positive DC transmission line 21 and a negative DC transmission line 22.

[0091] The inverter side AC simulation system 3 includes: an inverter side AC filter 33, an inverter side converter transformer 31, and an inverter side converter valve 32. The inverter side converter valve 32 is used to convert the DC power transmitted from the DC power transmission simulation line into AC power, the inverter side converter transformer 31 is used to transform the AC power transmitted from the inverter side converter valve 32, obtain the target AC power and output it, and the inverter side AC filter 33 is used to filter the inverter side.

[0092] In addition, the high-voltage direct current transmission system simulation model further includes a direct current filter 23 for filtering the direct current on the direct current transmission simulation line. One end of the rectifier-side converter valve 12 and the inverter-side converter valve 32 are both connected to the ground of the direct current transmission line.

[0093] The target sampling point is specifically located at the beginning or end of the positive DC transmission line 21 of the DC transmission simulation line, or at the beginning or end of the negative DC transmission line 22 .

[0094] In one embodiment, the target decomposition algorithm is a CEEMDAN decomposition algorithm.

[0095] Specifically, the Chinese name of CEEMDAN (Complete Ensemble Empirical Mode Decomposition with Adaptive Noise) is complete ensemble empirical mode decomposition with adaptive noise. This method is not improved on the CEEMD method, but improved on the basis of EMD. At the same time, it borrows the idea of ​​adding Gaussian noise and offsetting noise by multiple superposition and averaging in the EEMD method.

[0096] Step S200 specifically includes: using the CEEMDAN decomposition algorithm to decompose the simulated fault voltage signal data obtained in each simulation to obtain corresponding multiple components.

[0097] Step S600 specifically includes: decomposing the target fault data using the CEEMDAN decomposition algorithm to obtain corresponding target components.

[0098] The complete ensemble empirical mode decomposition method (CEEMDAN) is a signal decomposition method improved on the basis of empirical mode decomposition (EMD) and ensemble empirical mode decomposition (EEMD). CEEMDAN decomposition can solve the mode aliasing phenomenon existing in the EMD decomposition method, and overcome the defects of the incompleteness of the EEMD decomposition method and the low computational efficiency caused by increasing the number of integrations to reduce the reconstruction error. Moreover, the CEEMDAN decomposition method adds adaptive white noise at each stage of decomposition, calculates the unique residual signal to obtain each modal component, and its decomposition process is complete. It adaptively determines the number of decompositions according to the complexity of the four fault signals until the residual signal is a monotonic function.

[0099] CEEMDAN decomposes a set of simulation fault data and obtains:

[0100]

[0101] Among them, IMF 1 (t) is the high frequency component, imf j (t) is the intrinsic modal component, and the value of j is 2,...,N; where N is the number of components obtained after the same fault signal is decomposed, and N is a positive integer greater than or equal to 2, R 1 (t) is the residual term. The high frequency component and the natural mode component are the components corresponding to the simulated fault data.

[0102] The complete ensemble empirical mode decomposition (CEEMDAN) method is used to process fault signals, which solves the mode aliasing phenomenon of general decomposition methods.

[0103] In one embodiment, different fault types are simulated with different fault point locations set on the high voltage direct current transmission system simulation model.

[0104] Specifically, refer to Figure 2 When a DC line grounding fault is set at point A on the positive DC line of the HVDC transmission system simulation model, other parts of the HVDC transmission system simulation model remain normal and only point A has a DC line grounding fault.

[0105] Similarly, if Figure 2 The inverter side AC system three-phase transmission line B is set up with single-phase grounding fault, two-phase short-circuit grounding fault and three-phase grounding fault. The line at B is a single-line diagram. The specific settings of the single-phase grounding, two-phase short-circuit grounding and three-phase grounding faults of the inverter side AC system three-phase transmission line at the fault point at B are as follows: Figure 3As shown, when one of the faults is set, the other parts of the system remain normal. Among them, line a, line b, and line c are the A, B, and C phases of the three-phase transmission line.

[0106] In one embodiment, when a DC line fault occurs on a DC transmission simulation line of a simulation model of a high-voltage DC transmission system, the fault point set in the simulation model of the high-voltage DC transmission system is located on a positive DC transmission line or a negative DC transmission line in the DC transmission simulation line.

[0107] In one embodiment, in step S300, a corresponding fault feature vector is constructed according to multiple components corresponding to the same set of simulated fault data, including:

[0108] Calculate the sample entropy value corresponding to each component;

[0109] According to multiple sample entropy values ​​corresponding to the same set of simulated fault data, a corresponding fault feature vector is constructed;

[0110] or,

[0111] According to multiple components and sample entropy values ​​corresponding to the same set of simulated fault data, a corresponding fault feature vector is constructed.

[0112] Specifically, all acquired groups of simulated fault data are decomposed by CEEMDAN to obtain multiple components, and the sample entropy value of each component is calculated, and multiple sample entropy values ​​corresponding to the same group of simulated fault data are constructed into a fault feature vector, that is, a fault feature vector includes multiple sample entropy values.

[0113] Alternatively, the components and the corresponding sample entropy values ​​are concatenated together to form a fault feature vector corresponding to the simulated fault data. Both the components and the sample entropy values ​​are features that reflect the characteristics of the fault. The combination of the two into a fault feature vector enables the HVDC system fault diagnosis model to learn more features of different fault types, thereby enhancing the accuracy of the HVDC system fault diagnosis model.

[0114] In one embodiment, the pre-trained HVDC system fault diagnosis model is constructed based on a parallel convolutional neural network and a bidirectional long short-term memory neural network.

[0115] Specifically, the fault diagnosis model of the HVDC transmission system is based on the PCNN-BiLSTM structure of deep learning, which is mainly constructed by a parallel convolutional neural network (PCNN) and a bidirectional long short-term memory neural network (BiLSTM). The PCNN structure includes two parallel convolutional neural networks (CNN); the input fault feature vector is subjected to a deeper feature extraction in the PCNN structure. The two parallel different CNN structures are complementary to each other in the extraction of fault features, and the extraction forms of signal feature information are more diverse, which increases the available information dimension and helps to improve the fault diagnosis performance of the model; at the same time, the fault diagnosis model of the HVDC transmission system adopts a bidirectional long short-term memory neural network (BiLSTM), which can not only make full use of the temporal characteristics of the simulated fault data, but also take into account the future information and historical information of the fault data in a bidirectional manner, increase the available information of the fault diagnosis model, reduce the misjudgment rate of fault diagnosis, and improve the accuracy of fault diagnosis.

[0116] The use of parallel convolutional neural networks (PCNN) can extract deeper and more diverse fault information, and the use of bidirectional long short-term memory neural networks (BiLSTM) to identify faults expands the dimension of available fault information. This method effectively solves the problems of manually extracting fault information and manually determining model structure parameters in traditional fault identification methods, and can accurately distinguish different fault types.

[0117] In a specific embodiment, model training can be performed based on the Pelican Optimization Algorithm (POA) to optimize the model parameters of the high-voltage direct current transmission system fault diagnosis model, solve the appropriate PCNN convolution kernel size and number, pooling layer size, BiLSTM neuron number, learning rate and other parameters, and obtain the optimal fault diagnosis model.

[0118] The Pelican Optimization Algorithm (POA) is used to optimize the diagnostic model, which reduces the process of manually determining network parameters, reduces the randomness of diagnostic results, and improves the performance of the model.

[0119] The present application builds a simulation model of the Yongfu weak-receiving end high-voltage direct current transmission system through the PSCAD / EMTDC platform, sets different faults in the simulation model of the high-voltage direct current transmission system to obtain electrical experimental data reflecting different fault types, namely, simulated fault data; performs CEEMDAN decomposition on the obtained simulated fault data to obtain different components, reduces data complexity to obtain more features, and calculates the sample entropy value of each component to construct it into a fault feature vector reflecting the fault; then constructs a PCNN-BiLSTM high-voltage direct current transmission system fault diagnosis model based on the Pelican Optimization Algorithm (POA) optimization, determines the network parameters of the PCNN-BiLSTM high-voltage direct current transmission system fault diagnosis model through POA, and finally uses the optimized PCNN-BiLSTM high-voltage direct current transmission system fault diagnosis model to diagnose and identify the fault types of the actually operating high-voltage direct current transmission system.

[0120] Figure 4 This is a structural block diagram of a device for diagnosing a power transmission system fault in an embodiment of the present application. The device for diagnosing a high-voltage direct current power transmission system fault includes:

[0121] The first data acquisition module 100 is used to collect simulated fault data when different types of faults occur in multiple simulations of the high-voltage direct current transmission system simulation model, wherein the simulated fault data includes simulated fault voltage data of the target sampling point, and the different types of faults include a DC line fault and a single-phase grounding fault, a two-phase short-circuit grounding fault, and a three-phase grounding fault of the inverter-side AC system;

[0122] A decomposition module 200 is used to decompose the simulated fault data obtained from each simulation using a target decomposition algorithm to obtain corresponding multiple components;

[0123] The sample construction module 300 is used to construct a corresponding fault feature vector according to multiple different components corresponding to the same set of simulated fault data, and to construct a sample set according to multiple groups of different fault feature vectors corresponding to different fault types and corresponding known fault types;

[0124] A model training module 400 is used to train a pre-trained high-voltage direct current power transmission system fault diagnosis model using a sample set to obtain a trained high-voltage direct current power transmission system fault diagnosis model;

[0125] The second data acquisition module 500 is used to acquire target fault data collected at a target detection position when a fault occurs in the actually operating high voltage direct current transmission system, wherein the target detection position is a position in the actually operating high voltage direct current transmission system corresponding to a target sampling point in the high voltage direct current transmission system simulation model;

[0126] The decomposition module 200 is further used to decompose the target fault data using a target decomposition algorithm to obtain corresponding target components;

[0127] The diagnosis module 700 is used to form a target fault feature vector according to target components and input it into a trained high-voltage direct current transmission system fault diagnosis model to obtain the fault type of the fault occurring in the actually operating high-voltage direct current transmission system.

[0128] Figure 5 FIG. 1 shows an internal structure diagram of a computer device in an embodiment. The computer device may be a terminal or a server. Figure 5 As shown, the computer device includes a processor, a memory and a network interface connected via a system bus. The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system and may also store a computer program. When the computer program is executed by the processor, the processor may implement each step in the above method embodiment. The internal memory may also store a computer program. When the computer program is executed by the processor, the processor may implement each step in the above method embodiment. Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0129] In one embodiment, a computer device is proposed, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the following steps: collecting simulated fault data when different types of faults occur in multiple simulations of a high-voltage direct current transmission system simulation model, wherein the simulated fault data includes simulated fault voltage data of a target sampling point, and different types of faults include a DC line fault and a single-phase grounding fault, a two-phase short-circuit grounding fault, and a three-phase grounding fault of an inverter-side AC system; decomposing the simulated fault data obtained from each simulation using a target decomposition algorithm to obtain corresponding multiple components; and according to multiple different components corresponding to the same set of simulated fault data, Construct corresponding fault feature vectors, and construct a sample set according to multiple groups of different fault feature vectors corresponding to different fault types and corresponding known fault types; use the sample set to train a pre-trained high-voltage direct current transmission system fault diagnosis model to obtain a trained high-voltage direct current transmission system fault diagnosis model; obtain target fault data collected at a target detection position when a fault occurs in an actually operating high-voltage direct current transmission system; use a target decomposition algorithm to decompose the target fault data to obtain corresponding target components; form a target fault feature vector according to the target components and input it into the trained high-voltage direct current transmission system fault diagnosis model to obtain the fault type of the fault that occurs in the actually operating high-voltage direct current transmission system.

[0130] In one embodiment, a computer-readable storage medium is proposed, which stores a computer program. When the computer program is executed by a processor, the processor performs the following steps: collecting simulated fault data when different types of faults occur in multiple simulations of a high-voltage direct current transmission system simulation model, wherein the simulated fault data includes simulated fault voltage data of a target sampling point, and different types of faults include a DC line fault and a single-phase grounding fault, a two-phase short-circuit grounding fault, and a three-phase grounding fault of an inverter-side AC system; decomposing the simulated fault data obtained from each simulation using a target decomposition algorithm to obtain corresponding multiple components; constructing corresponding Fault feature vectors, construct a sample set according to multiple groups of different fault feature vectors corresponding to different fault types and the corresponding known fault types; use the sample set to train a pre-trained high-voltage direct current transmission system fault diagnosis model to obtain a trained high-voltage direct current transmission system fault diagnosis model; obtain target fault data collected at a target detection position when a fault occurs in an actually operating high-voltage direct current transmission system; use a target decomposition algorithm to decompose the target fault data to obtain corresponding target components; form a target fault feature vector according to the target components and input it into the trained high-voltage direct current transmission system fault diagnosis model to obtain the fault type of the fault that occurs in the actually operating high-voltage direct current transmission system.

[0131] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in the present application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0132] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0133] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for diagnosing a power transmission system fault, characterized in that: The method comprises: Collecting simulated fault data when different types of faults occur in multiple simulations of a high-voltage direct current transmission system simulation model, wherein the simulated fault data includes simulated fault voltage data of a target sampling point, and the different types of faults include a DC line fault and a single-phase grounding fault, a two-phase short-circuit grounding fault, and a three-phase grounding fault of an inverter-side AC system; Decomposing the simulated fault data obtained from each simulation by using a target decomposition algorithm to obtain corresponding multiple components; According to multiple components corresponding to the same set of simulated fault data, a corresponding fault feature vector is constructed, and according to multiple groups of different fault feature vectors corresponding to different fault types and corresponding known fault types, a sample set is constructed; Using the sample set to train a pre-trained high-voltage direct current power transmission system fault diagnosis model to obtain a trained high-voltage direct current power transmission system fault diagnosis model; Acquire target fault data collected at a target detection location when a fault occurs in an actually operating high voltage direct current transmission system; Decomposing the target fault data using the target decomposition algorithm to obtain corresponding target components; A target fault feature vector is formed according to the target components and input into the trained high-voltage direct current power transmission system fault diagnosis model to obtain the fault type of the fault occurring in the actually operating high-voltage direct current power transmission system.

2. The method according to claim 1, characterized in that The collecting of simulation fault data when different types of faults occur in multiple simulations of the high voltage direct current transmission system simulation model includes: Collect simulated fault voltage data of a target sampling point when a DC line fault occurs on a DC transmission simulation line in each simulation of the high voltage DC transmission system simulation model in a first preset number of times, and obtain simulated fault data corresponding to each DC line fault, wherein at least one of a fault distance and a resistance value of a transition resistance in different rounds of simulation in the first preset number of times of simulation is different; Collect simulated fault voltage data of a target sampling point when a single-phase grounding fault occurs in the inverter-side AC simulation system of each high-voltage direct current transmission system simulation model in the second preset number of simulations, and obtain simulated fault data corresponding to each single-phase grounding fault of the inverter-side AC system, wherein the resistance values ​​of the transition resistors in different rounds of simulation in the second preset number of simulations are different; Collect simulated fault voltage data of a target sampling point when a two-phase short-circuit grounding fault occurs in the inverter-side AC simulation system of each high-voltage direct current transmission system simulation model in the third preset number of simulations, and obtain simulated fault data corresponding to each two-phase short-circuit grounding fault of the inverter-side AC system, wherein the resistance values ​​of the transition resistors in different rounds of simulation in the third preset number of simulations are different; Collect simulated fault voltage data of the target sampling point when a three-phase grounding fault occurs in the inverter-side AC simulation system of each high-voltage direct current transmission system simulation model in the fourth preset number of simulations, and obtain a simulated fault signal corresponding to each three-phase grounding fault of the inverter-side AC system, wherein the resistance values ​​of the transition resistances simulated in different rounds of the fourth preset number of simulations are different.

3. The method according to claim 2, characterized in that The simulation model of the simulated high-voltage direct current transmission system includes a rectifier-side AC simulation system, a DC transmission simulation line and an inverter-side AC simulation system connected in sequence; The target sampling point is located on the DC power transmission simulation line.

4. The method according to claim 2, characterized in that: When simulating different fault types, the locations of the obstacle points set on the high-voltage direct current transmission system simulation model are different; When a DC line fault occurs on the DC transmission simulation line of the simulation model of the high-voltage DC transmission system, the fault point set in the simulation model of the high-voltage DC transmission system is located on the positive DC transmission line or the negative DC transmission line in the DC transmission simulation line.

5. The method according to claim 1, characterized in that The method of constructing a corresponding fault feature vector according to multiple components corresponding to the same set of simulated fault data includes: Calculate the sample entropy value corresponding to each component; According to multiple sample entropy values ​​corresponding to the same set of simulated fault data, a corresponding fault feature vector is constructed; or, According to multiple components and sample entropy values ​​corresponding to the same set of simulated fault data, a corresponding fault feature vector is constructed.

6. The method according to claim 1, characterized in that The pre-trained high-voltage direct current transmission system fault diagnosis model is constructed based on a parallel convolutional neural network and a bidirectional long short-term memory neural network.

7. The method according to claim 1, characterized in that The target decomposition algorithm is the CEEMDAN decomposition algorithm.

8. A diagnostic device for a high voltage direct current transmission system fault, characterized in that: The device comprises: A first data acquisition module is used to collect simulated fault data when different types of faults occur in multiple simulations of a high-voltage direct current transmission system simulation model, wherein the simulated fault data includes simulated fault voltage data of a target sampling point, and the different types of faults include a DC line fault and a single-phase grounding fault, a two-phase short-circuit grounding fault, and a three-phase grounding fault of an inverter-side AC system; A decomposition module, used for decomposing the simulation fault data obtained from each simulation by using a target decomposition algorithm to obtain corresponding multiple components; A sample construction module is used to construct a corresponding fault feature vector according to multiple components corresponding to the same set of simulated fault data, and to construct a sample set according to multiple groups of different fault feature vectors corresponding to different fault types and corresponding known fault types; A model training module, used to train a pre-trained high-voltage direct current power transmission system fault diagnosis model using the sample set to obtain a trained high-voltage direct current power transmission system fault diagnosis model; The second data acquisition module is used to acquire target fault data collected at a target detection position when a fault occurs in the actually operating high voltage direct current transmission system; The decomposition module is further used to decompose the target fault data using the target decomposition algorithm to obtain corresponding target components; The diagnosis module is used to form a target fault feature vector according to the target components and input it into the trained high-voltage direct current power transmission system fault diagnosis model to obtain the fault type of the fault occurring in the actually operating high-voltage direct current power transmission system.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 7.

10. A computer device comprising a memory and a processor, characterized in that: The memory stores a computer program, and when the computer program is executed by the processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 7.

Images