Power distribution network resonance grounding fault identification method, system and device, and storage medium
The zero-sequence current waveform is analyzed through VMD and MFDFA algorithms, and combined with the SVM classifier, the problem of low-resistance and high-resistance grounding fault identification in the distribution network is solved, and fault identification and distinction with high accuracy is achieved.
Patent Information
- Application Number
- CN202411479214.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-23
- Publication Date
- 2024-11-22
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The prior art is difficult to effectively identify and detect low-resistance grounding and high-resistance grounding faults in the resonant grounding of the distribution network. Especially high-resistance grounding faults are more difficult to detect due to the small fault current.
The VMD algorithm combined with MFDFA multi-fractal detrend fluctuation analysis method is used to analyze the zero-sequence current waveform. The eigenvectors are constructed through different eigenvalues of fractal parameters, and fault identification is used by SVM classifiers to realize automatic low-resistance and high-resistance faults. Identify.
It improves the accuracy of fault identification, can be highly adaptable under different simulation conditions, accurately distinguishes low-resistance and high-resistance faults, and improves the detection ability of resonant grounding faults in the distribution network.
Smart Images

Figure CN119001338A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of distribution network fault detection, and in particular relates to a distribution network resonant grounding fault identification method, system, equipment and storage medium. Background Art
[0002] There are low-resistance grounding faults and high-resistance grounding faults in the resonant grounding system of the distribution network, and the low-resistance grounding faults and high-resistance grounding faults have different characteristics.
[0003] Low-resistance grounding systems are relatively easy to detect because the grounding transition resistance is small and the current rises quickly when a fault occurs.
[0004] Compared with low-resistance grounding faults, high-resistance grounding faults are more difficult to detect due to their large grounding transition resistance, small fault current, and unstable fault point.
[0005] How to realize the automatic identification and detection of low-resistance grounding and high-resistance grounding faults in the resonant grounding of distribution networks has always been a major problem that has troubled power workers. Summary of the invention
[0006] The present invention proposes a method, system, device and storage medium for identifying resonant grounding faults in a distribution network. The VMD algorithm is combined with the MFDFA multi-fractal detrended fluctuation analysis method to analyze the zero-sequence current waveform under different states of high-resistance faults and low-resistance faults. The method has strong adaptability to different simulation conditions and the recognition result has high accuracy.
[0007] The embodiments of the present invention are preferably implemented by the following scheme: In a first aspect, the present invention provides a method for identifying a resonant grounding fault in a distribution network, comprising dividing the transient process of a resonant grounding system fault into a low-resistance underdamped process, an overdamped process and a high-resistance underdamped process according to different resonant frequencies; solving the fractal parameters of the low-resistance underdamped process, the overdamped process and the high-resistance underdamped process; extracting fault feature quantities of the low-resistance underdamped process, the overdamped process and the high-resistance underdamped process according to the fractal parameters; combining the fault feature quantities into a feature vector of a fault signal, extracting feature vectors from historical resonant grounding fault data of the power grid to form a training sample set; training an SVM classifier model based on the training sample set; and using the trained model to identify faults.
[0008] Furthermore, Solving the fractal parameters of the low-resistance underdamped process, the overdamped process and the high-resistance underdamped process includes: using the MFDFA algorithm to obtain the fractal parameters; using the VMD algorithm to decompose the fractal parameters to obtain signal residuals; substituting the signal residuals into the MFDFA algorithm to calculate the optimized fractal parameters.
[0009] Furthermore, According to the fractal parameters, the fault characteristics of the low-resistance under-damping process, the over-damping process and the high-resistance under-damping process are extracted, including: extracting the maximum value, the minimum value, and the difference and standard deviation between the maximum value and the minimum value of the optimized fractal parameters as the fault characteristics.
[0010] Furthermore, The SVM classifier has a binary tree structure.
[0011] In the second aspect, the present invention provides a distribution network grounding fault identification system, including a transient process classification module, a fractal parameter solving module, a fault feature extraction module, a training sample generation module and a fault analysis module; the transient process classification module is used to divide the transient process of the resonant grounding system fault into a low-resistance underdamped process, an overdamped process and a high-resistance underdamped process according to different resonant frequencies; the fractal parameter solving module solves the fractal parameters of the low-resistance underdamped process, the overdamped process and the high-resistance underdamped process; the fault feature extraction module extracts the fault feature quantities of the low-resistance underdamped process, the overdamped process and the high-resistance underdamped process according to the fractal parameters; the training sample generation module combines the fault feature quantities into a feature vector of a fault signal, and extracts feature vectors from the historical resonant grounding fault data of the power grid to form a training sample set; the fault analysis module trains the SVM classifier and identifies faults based on the training sample set.
[0012] In a third aspect, the present invention provides an electronic device, comprising: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement a method for identifying a resonant grounding fault in a distribution network.
[0013] In a fourth aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for identifying a resonant grounding fault in an existing power distribution network.
[0014] In a fifth aspect, the present invention further provides a computer program, which, when executed by a processor, implements a method for identifying a resonant grounding fault in an existing power distribution network.
[0015] The present invention uses multi-fractal detrended fluctuation analysis (MFDFA) to analyze the zero-sequence current waveform under different states of high-resistance fault and low-resistance fault. Since MFDFA can finely characterize the local characteristics of the signal and amplify the local scale behavior of the signal, when a high-resistance fault and a low-resistance fault occur, their fractal parameter sizes are significantly different. Four eigenvalues of the fractal parameters are selected to construct a feature vector, and then a support vector machine (SVM) that can effectively solve small sample classification is used to input the feature vector into a multi-level support vector machine to achieve effective identification of high-resistance faults.
[0016] The technical effects of the embodiments of the present invention are as follows: Firstly, by analyzing the transient process of low-resistance fault and high-resistance fault under different states, it is found that the fault transient can be divided into three states according to the different transition resistances: low-resistance underdamping, overdamping and high-resistance underdamping. Secondly, the zero-sequence current of the arc-extinguishing coil under different states is analyzed by using variational mode decomposition VMD combined with multi-fractal detrending method MFDFA to determine the fractal parameters. The fractal parameter curve under low-resistance underdamping fault is a straight line with a small range of variation; the fractal parameter curve under high-resistance underdamping fault is inverse tangent distribution due to the beat frequency oscillation, and the range of variation is large; while the fractal parameter under overdamping fault is between the two. Finally, the four eigenvalues of the fractal parameter are selected to form a feature vector, and the feature vector is input into a multi-level support vector machine (SVM) to realize fault identification.
[0017] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0019] Figure 1 A schematic diagram of a flow chart of a method for identifying ground faults in a distribution network according to an embodiment of the present invention; Figure 2 is a transient equivalent circuit diagram of a low-resistance grounding fault in an embodiment of the present invention; Figure 3 It is a transient equivalent circuit diagram of a high-resistance grounding fault in an embodiment of the present invention; Figure 4 This is a schematic diagram of a binary tree structure of an SVM classifier in an embodiment of the present invention; Figure 5 A schematic diagram of the structure of a distribution network grounding fault identification system according to an embodiment of the present invention; Figure 6 The figure is a schematic diagram of the structure of an electronic device in an embodiment of the present invention. DETAILED DESCRIPTION
[0020] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0021] The disclosure below provides many different embodiments or examples to implement different structures of the present invention. In order to simplify the disclosure of the present invention, the parts and settings of specific examples are described below. Of course, they are only examples, and the purpose is not to limit the present invention. In addition, the present invention can repeat reference numbers and / or letters in different examples. This repetition is for the purpose of simplification and clarity, and does not itself indicate the relationship between the various embodiments and / or settings discussed.
[0022] Attached Figure 1 A schematic flow chart of a method for identifying a resonant grounding fault in a distribution network according to an embodiment of the present invention is shown, and this embodiment will be described in detail below in conjunction with the accompanying drawings.
[0023] Step 1: According to the different resonant frequencies, the transient process of the resonant ground fault is divided into three processes: low-resistance underdamping, overdamping and high-resistance underdamping. This step is the prior art.
[0024] When a single-phase grounding fault occurs in the resonant grounding system of the distribution network, the different transition resistances of the grounding point correspond to different states of the fault transient process. The following is an analysis of the transient process during low-resistance and high-resistance faults in the prior art.
[0025] 1. Analysis of transient process of low resistance fault When the transition resistance of the grounding point is small, the zero-sequence current flowing through the line is large. At this time, the transient main resonant frequency of the resonant grounding system is relatively high. Taking the third harmonic as an example, the inductive reactance of the arc suppression coil will be 9 times greater than the capacitive reactance of the resonant grounding system to the ground in parallel with it, and the effect of the arc suppression coil can be ignored.
[0026] The transient resonance process of the resonant grounding system is composed of the series resonance of the line inductance and the capacitance to the ground. The transient equivalent circuit of the low-resistance grounding fault is as follows: Figure 2 shown.
[0027] Figure 2 In the figure, the second-order linear nonhomogeneous differential equation of the transient equivalent circuit of the low-resistance grounding fault is: (1), in: It is the virtual power supply at the fault point. is the virtual power amplitude; L It is the sum of the zero-mode inductance, first-mode inductance and second-mode inductance of the line from the fault point to the busbar; the equivalent resistance R It is the sum of the zero-mode resistance, first-mode resistance, second-mode resistance and three times the transition resistance of the line from the fault point to the busbar; C ∑ It is the sum of the capacitance of all lines to ground in the system; is the bus zero-sequence voltage; is the bus zero-sequence power frequency.
[0028] The characteristic root of equation (1) is: (2).
[0029] In formula (2), P1 and p2 represent characteristic roots.
[0030] When the equivalent resistance R satisfy: When , equation (1) has two conjugate complex roots with negative real parts, and the transient process of the resonant grounding system is underdamped. In a typical resonant grounding 10kV distribution network, considering the general situation of the resonant grounding system capacitance current and line length, the critical resistance value between underdamping and overdamping in a low-resistance grounding fault is 9Ω~208Ω.
[0031] The transient component of the zero-sequence current at the low-resistance grounding fault point is: (3).
[0032] The resonant frequency of the transient component of a low-resistance ground fault is: (4).
[0033] The magnitude of the resonant frequency is affected by factors such as the grounding resistance at the fault point, line inductance, total system-to-ground distributed capacitance, and line length from the fault point to the busbar. It decreases with the increase of line length and system-to-ground capacitance. Usually, the resonant frequency ranges from 350 to 2190 Hz, which is much larger than the fundamental frequency.
[0034] When the equivalent resistance satisfies: When , equation (1) has two unequal negative real roots. At this time, the system is in an overdamped state, the transient process is a non-oscillating process, and the transient component of the zero-sequence current at the fault point is: (5).
[0035] In formula 3: , is the characteristic vector of zero-sequence current at the fault point; A 1, A 2 is for R , L , C ∑ The size of the correlation coefficient.
[0036] At this time, the transient zero-sequence current is two attenuated DC components, the corresponding resonant frequency is 0, and only a small part of the fault transient is in an over-damped state.
[0037] 2. Analysis of transient process of high resistance fault When the transition resistance of the grounding point is large, the zero-sequence current flowing through the resonant grounding system is small, and the charging and discharging speed of the zero-sequence capacitance on the line is slow, which makes the transient resonant frequency of the resonant grounding system low. At this time, the line inductance L The inductive reactance is small and its influence can be ignored. The resonance process of the resonant grounding system is composed of the parallel resonance of the ground capacitance and the equivalent inductance of the arc suppression coil. Its fault transient equivalent circuit is as follows Figure 3 As shown. Among them, Lp It is the equivalent inductance of arc suppression coil (3 times the inductance of arc suppression coil).
[0038] Figure 3 The second-order linear nonhomogeneous differential equation of the transient equivalent circuit of medium-high resistance ground fault is: (6).
[0039] In formula (6): i Lp is the current flowing through the inductor Lp of current.
[0040] The characteristic root of equation (6) is: (7).
[0041] When the equivalent resistance R satisfy: When , the transient process of high-resistance grounding fault is in an over-damped state. At this time, the system has two unequal negative real roots and the resonant frequency is 0.
[0042] Therefore, the equivalent resistance R Should meet: When the resonant grounding system is in an over-damped state, no matter whether the low-resistance model or the high-resistance model is used, the transient electrical quantity of the fault is the superposition of two attenuated DC components, and there is no oscillation process. Considering the general distribution of the ground capacitance and arc suppression coil of the resonant grounding system, the equivalent resistance value in the over-damped state is about 100 ohms.
[0043] When the equivalent resistance R satisfy: When the resonant grounding system is in an under-damped state, the transient electrical quantity of the fault will oscillate. In a general 10kV distribution network resonant grounding system, the critical grounding resistance value ranges from 29Ω to 300Ω. R The value that needs to be met is around 100 ohms, and is usually tens of ohms. However, the transition resistance of a high-resistance fault can reach thousands of ohms, so most high-resistance faults are in an under-damped state.
[0044] Solving equation (6), we can get the expression of zero-sequence current at the fault point: (8), In formula (8): A 3. A 4 is for R , L p , C ∑ The coefficient related to the size of δ is the attenuation factor, which determines the attenuation speed of the electrical quantity; is the vector corresponding to the attenuation factor.
[0045] The transient zero-sequence current of the fault is two attenuated sinusoidal quantities, and the transient resonant frequency of the resonant grounding system is for: (9), In the formula v is the detuning degree of the arc suppression coil, which is -0.08 in the embodiment of the present invention. Obviously, the system resonant frequency increases with the increase of the fault grounding resistance. According to the general values of the ground capacitance and the transition resistance, the upper limit of the resonant frequency can be obtained as: (10), The corresponding frequency is 52Hz.
[0046] It can be seen that in the under-damped state of a high-resistance grounding fault, as the transition resistance continues to increase, the value of the resonant frequency gradually increases, but the value of the resonant frequency will not increase infinitely, and the maximum resonant frequency is only slightly higher than the power frequency. When the transient resonant frequency is greater than 47Hz, the equivalent resistance R The conditions to be met are: (11).
[0047] Since the capacitive reactance of the ground capacitor in the system is generally hundreds of ohms, the transition resistance can have a large range of values, that is, most high-resistance faults in the underdamped state will show beat frequency phenomenon.
[0048] In summary, according to the different resonant frequencies (corresponding to different equivalent resistances R ), the transient process of resonant grounding fault can be divided into three states: low-resistance underdamping, overdamping and high-resistance underdamping. Among them, the resonant frequency of low-resistance underdamping is relatively high, generally ranging from 350 to 2190Hz, and the corresponding transition resistance value is very small; the overdamped state can be simulated by either a low-resistance model or a high-resistance model, the resonant frequency of this state is 0, and the transition resistance value is about 100 ohms; the highest resonant frequency of the high-resistance underdamping state is close to the power frequency, and most faults will have a beat frequency phenomenon, and the transition resistance value can reach thousands of ohms.
[0049] Step 2: Solve the fractal parameters of the three states: low resistance underdamping, overdamping and high resistance underdamping.
[0050] 1. Use multi-fractal detrending volatility analysis method Multifractal detrended fluctuation analysis (MFDFA) is a generalized definition of detrended fluctuation analysis (DFA). MFDFA can weaken or eliminate the unknown trend components in the original signal, effectively extract the fractal characteristics of the original signal, and finely characterize the local scale characteristics of the signal. It has good superiority in the multifractal feature analysis of non-stationary signals. The specific analysis steps are as follows: Based on formula (8), the zero-sequence current time series of the fault point is constructed as follows: x i , i =1, 2, ..., N}, solve for the cumulative deviation of the zero-sequence current mean Y ( i ): (12), In formula (12): x i is the zero-sequence current of the fault point to be detected; is the mean value of the zero-sequence current time series; N , M is a natural number, M < N .
[0051] The cumulative deviation of the constructed zero-sequence current time series from the mean Y(i) With fixed scale k Divide into N k Segment sequence. When the data length N Not divisible k In order to ensure that the data length can be fully utilized, it can be divided from the positive and negative directions. First, the time series to be detected { xi , i =1, 2, ..., N} from the positive direction into non-overlapping N k The segment sequence is then divided into non-overlapping N k A total of 2 N k Segment data: (13), In formula (13), k is the time series scale.
[0052] The least square method is used to perform n ( n =1, 2, 3, ...) order polynomial fitting, n The larger the value, the better the detrending effect. v ( j ) is the v The least squares of subsequences n Order fitting polynomial: (14), In formula (14): a k ( j = 1, 2, …, k ) is the fitting polynomial, n is the order of the quasi-polynomial. The local covariance is obtained function: (15).
[0053] After detrending, 2 N k subsequences, find The average value of all 2 N k Subsequence by q The order wave function is obtained by the following equation: (16), In formula (16): F q (k) Is the data length k and fractal order q The wave function of k There is a stable power law relationship, namely: (17), Solve the zero-sequence current fractal parameters H(q) : (18).
[0054] 2. VMD-based MFDFA method VMD Decomposition Algorithm VMD is an adaptive signal decomposition method that uses a non-recursive method to decompose the signal into several IMFs with different center frequencies according to the predetermined number of modes and penalty factors.
[0055] VMD transforms the signal into a restricted variational problem, which is expressed as follows: (19), In formula (19): is a real signal, G is the number of decomposed signals, is the modal function, is the center frequency.
[0056] By introducing the penalty factor With Lagrange multipliers Transform the above problem into an unrestricted variation problem: (20), The decomposed signal residual can be obtained by distributed iterative updating of the penalty factor and the Lagrange multiplier.
[0057] In order to optimize the detrending problem in the MFDFA algorithm, the VMD algorithm shown in Eqs. (19)-(20) is embedded in the MFDFA polynomial fitting process. The fractal parameters H(q) As the real signal in (19) , the signal residual after VMD decomposition is used to replace the time series of the original zero-sequence current at the fault point { x i , i =1, 2, ..., N}, the steps of formula (12)-(18) remain unchanged. It can be obtained: (twenty one), In formula (21): is the residual after VMD decomposition.
[0058] Based on formula (21), after calculation of formula (16)-formula (18), the optimized fractal parameter can be obtained: H(q)' .
[0059] Step 3: Extract fault characteristics of three states: low resistance underdamping, overdamping and high resistance underdamping.
[0060] In a resonant grounding system, when a low-resistance fault occurs, since the resistance of the transition resistor is only a few ohms, the decay time of the transient electrical quantity is very short, and the zero-sequence current waveform quickly returns to a sinusoidal wave. When a high-resistance grounding fault occurs, as the transition resistance increases, the decay rate of the electrical quantity slows down, because the transition resistance in the over-damped state is about 100 ohms. The transition resistance in the under-damped state reaches thousands of ohms, and there is a beat frequency phenomenon, so there is a significant difference in the duration of the transient process between the two. The zero-sequence current of the arc suppression coil in the three states of low-resistance underdamping, overdamping and high-resistance overdamping within two power frequency cycles after the fault occurs is sampled respectively. Since the degree of self-similarity of the three waveforms is significantly different, the optimized fractal parameters after multi-fractal detrending fluctuation analysis are used. H(q)' According to this difference, three fault states are distinguished: low resistance underdamping, overdamping and high resistance underdamping.
[0061] Due to the optimized fractal parameters of the signal H(q)' The amount of data is large. If it is directly classified, it will affect the calculation speed. H(q)' The maximum value of the curve H(q)' max , minimum H(q)' min , the difference between the maximum and minimum values △ H (q)' , Standard Deviation S td The four types of distribution parameters are combined into a characteristic vector of a fault signal: X={H(q)' max ,H(q)' min , △H(q)', S td}.
[0062] Step 4: Fault identification.
[0063] 1. Extract the feature vector X={ H(q)' max , H(q)' min ,△ H(q)' , S td}, forming a training sample set.
[0064] 2. Build a SVM classifier.
[0065] In order to identify the three fault states of low resistance underdamping, overdamping and high resistance underdamping, a two-level SVM classifier needs to be established. The SVM binary tree structure is as follows: Figure 4 shown.
[0066] 3. Kernel function (RBF) parameter optimization.
[0067] According to the SVM algorithm, for the feature vector X={ H(q)' max , H(q)' min ,△ H(q)' , S td}, there exists a hyperplane , so that the eigenvector X={ H(q)' max , H(q)' min ,△ H(q)' , S td Two types of data are distinguished in}, where is the normal vector of the optimal classification surface, b is the classification threshold. Nonlinear samples The optimization problem of (zero-sequence current of the fault point to be detected) data can be expressed as: (twenty two), In formula (22): is the error vector, is the relaxation factor, c is the penalty factor, represents the objective function, Indicates that the sample Functions that map to higher-dimensional linear spaces.
[0068] The Lagrange multiplier method is used to solve the optimization problem in equation (22), and the output function is obtained as follows: (twenty three), In formula (23): is the kernel function, is the optimal Lagrange multiplier, b * is the optimal threshold.
[0069] The expression of RBF kernel function is as follows: (twenty four), In formula (24): x is the feature vector of the sample to be tested, is the kernel parameter, which determines the complexity of the sample data distribution.
[0070] The output of the RBF kernel function is represented by -1, 0, and 1. -1 indicates that the fault type is a low-resistance under-damping fault, 0 indicates that the fault type is an over-damping fault, and 1 indicates that the fault type is a high-resistance under-damping fault.
[0071] Based on the same inventive concept, the present disclosure also proposes a high-resistance fault identification system for a resonant grounding system, including a transient process classification module, a fractal parameter solution module, a fault feature extraction module, a training sample generation module and a fault analysis module.
[0072] A transient process classification module is used to classify the transient process of the resonant grounding system fault into a low-resistance under-damping process, an over-damping process and a high-resistance under-damping process according to different resonant frequencies; Fractal parameter solving module, solving the fractal parameters of low-resistance underdamping process, overdamping process and high-resistance underdamping process; The fault feature extraction module extracts the fault feature quantities of the low-resistance under-damping process, the over-damping process and the high-resistance under-damping process according to the fractal parameters; The training sample generation module combines the fault feature quantities into a feature vector of a fault signal, and extracts feature vectors from the historical resonant grounding fault data of the power grid to form a training sample set; Fault analysis module, training SVM classifier and fault identification based on training sample set Based on the same inventive concept, the embodiment of the present disclosure further provides a computer-readable storage medium storing one or more programs, which, when executed, can implement the aforementioned high-resistance fault identification method for a resonant grounding system.
[0073] Based on the same inventive concept, Figure 6 As shown, the embodiment of the present disclosure also provides an electronic device, including a processor 301, a communication interface 302, a memory 303 and a communication bus 304. The processor, the communication interface and the memory communicate with each other via the communication bus. The processor is used to execute the program stored in the aforementioned computer-readable storage medium.
[0074] It should be noted that the electrical connection between the above-mentioned units does not necessarily mean the connection between lines. An indirect connection method can be applied to the embodiments of the present disclosure as long as the purpose of the present disclosure is achieved.
[0075] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a general hardware platform, and of course, by hardware. Based on this understanding, the above technical solution, in essence, or the part that contributes to the relevant technology, can be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiment.
[0076] It should be understood that the terms used herein are only for the purpose of describing specific example embodiments and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used herein may also be meant to include plural forms. The terms "include", "comprise", "contain", and "have" are inclusive, and therefore specify the existence of stated features, steps, operations, elements and / or parts, but do not exclude the existence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes, and operations described herein are not interpreted as necessarily requiring them to be performed in the specific order described or illustrated, unless the execution order is clearly indicated. It should also be understood that additional or alternative steps may be used.
[0077] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.
Claims
1. A method for identifying a resonant grounding fault in a distribution network, characterized in that: include: According to the different resonant frequencies, the transient process of the resonant grounding system fault is divided into low-resistance under-damping process, over-damping process and high-resistance under-damping process. Solve the fractal parameters of low-resistance underdamped process, overdamped process and high-resistance underdamped process; According to the fractal parameters, the fault characteristics of low-resistance under-damping process, over-damping process and high-resistance under-damping process are extracted; The fault feature quantities are combined into a feature vector of a fault signal, and the feature vectors are extracted from the historical resonant grounding fault data of the power grid to form a training sample set; The SVM classifier model is trained based on the training sample set; and the trained model is used for fault identification.
2. The method according to claim 1, characterized in that Solve the fractal parameters of low-resistance underdamped process, overdamped process and high-resistance underdamped process, including: Use the MFDFA algorithm to find the fractal parameters; The VMD algorithm is used to decompose the fractal parameters to obtain the signal residual; Substitute the signal residual into the MFDFA algorithm to calculate the optimized fractal parameters.
3. The method according to claim 2, characterized in that According to the fractal parameters, the fault characteristics of low-resistance under-damping process, over-damping process and high-resistance under-damping process are extracted, including: The maximum value, minimum value, difference between the maximum value and the minimum value and standard deviation of the optimized fractal parameters are extracted as fault characteristics.
4. The method according to claim 3, characterized in that The SVM classifier has a binary tree structure.
5. A distribution network resonant grounding fault identification system, characterized in that: It includes transient process classification module, fractal parameter solution module, fault feature extraction module, training sample generation module and fault analysis module; A transient process classification module is used to classify the transient process of the resonant grounding system fault into a low-resistance under-damping process, an over-damping process and a high-resistance under-damping process according to different resonant frequencies; Fractal parameter solving module, solving the fractal parameters of low-resistance underdamping process, overdamping process and high-resistance underdamping process; The fault feature extraction module extracts the fault feature quantities of the low-resistance under-damping process, the over-damping process and the high-resistance under-damping process according to the fractal parameters; The training sample generation module combines the fault feature quantities into a feature vector of a fault signal, and extracts feature vectors from the historical resonant grounding fault data of the power grid to form a training sample set; The fault analysis module trains the SVM classifier and identifies faults based on the training sample set.
6. An electronic device, characterized in that: include: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 4.
7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 4 is implemented.
8. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 4 are implemented.
Citation Information
Patent Citations
Sea wave height analysis method
CN104834811A
VMD equipment fault diagnosis method and system
CN112881045A
Feature extraction method and system for electroencephalogram signals
CN115186719A
Ground fault monitoring method, monitoring device and fault indicator of distribution line
CN116559724A
Adaptive fault phase selection method based on wavelet packet reconstruction energy of three-phase voltage signal
CN117849519A
