Method, system and device for positioning alternating current side fault of alternating current and direct current overline fault

Abstract

The invention relates to the technical field of alternating-current and direct-current power transmission, in particular to an alternating-current and direct-current overline fault alternating-current side fault positioning method, system and device, and the method comprises the steps: obtaining a three-phase alternating-current value in real time, and monitoring whether a direct-current bipolar network and a three-phase alternating-current network fluctuate or not; if the three-phase alternating current value changes, decomposing the three-phase alternating current value to obtain a decomposed low-frequency-band time-domain curve; establishing a fault positioning analysis model, performing stability analysis on the low-frequency-band time-domain curve, and calculating a stability index; and performing interval division on the low-frequency-band time-domain curve, traversing an interval division result based on the stability index, and judging a fault type occurring in each interval. According to the invention, the problem of rapid detection and positioning of the AC-side fault of the AC-DC overline fault is effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of AC / DC power transmission, and in particular to a method, system and device for locating an AC-side fault of an AC / DC cross-line fault. Background Art

[0002] With the accelerated construction of high-voltage direct current transmission projects, transmission corridors are becoming increasingly tight, and the power grid is gradually forming a mixed operation state of AC and DC lines. The crossing of AC and DC lines is relatively common. The new type of AC and DC lines installed on the same tower saves corridor land and construction costs, and is gradually being promoted and applied.

[0003] When AC and DC lines cross and are installed on the same tower, the distance between the AC and DC lines is relatively close, making cross-line faults more likely to occur. When AC and DC cross-line faults occur, the AC quantity will invade the DC system through the line contact point, and the DC quantity will also invade the AC system, causing asymmetric operation of the AC and DC systems, seriously damaging the line insulation and the safe and stable operation of the main equipment of the AC and DC stations.

[0004] The information disclosed in this background technology section is only intended to deepen the understanding of the overall background technology of the present disclosure, and should not be regarded as acknowledging or suggesting in any form that the information constitutes the prior art known to those skilled in the art. Summary of the invention

[0005] The present invention provides a method and system for locating an AC-side fault of an AC-DC cross-line fault, which can effectively solve the problems in the background technology.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is:

[0007] A method for locating an AC-side fault of an AC-DC cross-line fault, the method comprising:

[0008] Obtain three-phase AC current values ​​in real time and monitor whether there are fluctuations in the DC bipolar network and the three-phase AC network;

[0009] If there is a change, the three-phase AC current value is decomposed to obtain a low-frequency time domain curve after decomposition;

[0010] Establishing a fault location analysis model, performing a stability analysis on the low-frequency time domain curve, and calculating the stability index;

[0011] The low-frequency time domain curve is divided into intervals, and the interval division results are traversed based on the smoothness index to determine the type of fault occurring in each interval.

[0012] Furthermore, a fault location analysis model is established, including:

[0013] Analyzing the topological structure, operating status and current characteristics of the DC bipolar network and the three-phase AC network and establishing an initial spatial model;

[0014] Based on the initial spatial model, random fault simulation is performed to generate current fluctuation characteristics under different conditions;

[0015] Analyze the current fluctuation characteristics and sort out the trigger conditions, and generate a fault location analysis model based on the training of the trigger conditions;

[0016] The random fault simulation is performed again, model parameters and fault diagnosis algorithms are continuously optimized according to the fault simulation results, and the fault location analysis model is iteratively updated.

[0017] Furthermore, random fault simulation is performed, including:

[0018] Randomly select fault locations in DC bipolar networks and three-phase AC networks;

[0019] Randomly determine any fault type of a cross-line fault and a ground fault, and define different current fluctuation characteristics according to the fault type;

[0020] Setting different transition resistance parameters for each of the fault types, simulating different fault severities, and generating multiple current fluctuation modes;

[0021] According to the current fluctuation characteristics and the current fluctuation pattern, the fault location analysis model is preliminarily trained to perform fault location and type classification for subsequent fault scenarios.

[0022] Further, the three-phase alternating current value is decomposed, including:

[0023] Using wavelet transform to perform multi-layer decomposition on the three-phase AC current value, and extracting the decomposed signals of different frequency bands;

[0024] Extracting main fluctuation information and fault characteristics from the low-frequency time domain curve obtained by decomposition;

[0025] Analyze the high-frequency time domain curve, evaluate the detailed changes of the current signal, and assist in distinguishing high-frequency noise from fault signals in the three-phase AC network;

[0026] The stability analysis is performed based on the decomposition results of the low-frequency time domain curve and the high-frequency time domain curve in combination with the time characteristics of the three-phase alternating current value.

[0027] Further, extracting main fluctuation information from the decomposed low-frequency time domain curve includes:

[0028] The low-frequency time domain curve is multi-layered decomposed according to wavelet transform, and the low-level wavelet frequency band time domain image after the decomposition of the low-frequency time domain curve is obtained, and the time domain signal is defined as f(t). The wavelet transform is specifically as follows:

[0029]

[0030] Where W f (a,b) represents the transform coefficient of the signal f(t) at scale a and position b, is the mother wavelet function, which is transformed by the scale factor a and the translation factor b They are collectively called wavelets;

[0031] For discrete time series:

[0032]

[0033] The overall formula represents the function of the mother wavelet in the discrete time series after scale transformation j and translation transformation k. represents the sub-wavelet function after scale and translation transformation, 2 -j / 2 represents the scale factor, -k represents the translation factor;

[0034] Obtaining the main fluctuation information of the low-frequency time domain curve by wavelet transform, including the amplitude change, fluctuation period and frequency distribution of the signal;

[0035] The main fluctuation information of the low-frequency time domain curve is combined with the analysis result of the high-frequency time domain curve to determine the fault characteristics in the current signal.

[0036] Further, calculating the stability index includes:

[0037] Calculate the average value of the low-frequency time domain curve and calculate the stability index based on the main fluctuation information:

[0038]

[0039] Among them, S(I i ) represents the stability index, x i represents the i-th sample, represents the average value, and N represents the total number of data points;

[0040] According to the smoothness index, the fault type is judged by interval division, and the smoothness index of each interval is compared with a preset fault threshold to determine whether the fault condition is met.

[0041] Furthermore, the low-frequency time domain curve is divided into intervals, including:

[0042] According to the stability index of the low-frequency time domain curve, a preset threshold range is determined, and a plurality of intervals are set for dividing different fluctuation characteristics;

[0043] According to the change of the stability index, the low-frequency time domain curve is divided into intervals according to the threshold range, and a fluctuation pattern corresponding to each interval is generated;

[0044] Analyze each divided interval to determine whether the fluctuation characteristics of each interval meet the preset fault standard;

[0045] If the fluctuation within the interval exceeds the set threshold, the fault type occurring in each interval is determined based on the fluctuation pattern within the interval.

[0046] Furthermore, the fault type occurring in each interval is determined, including:

[0047] According to the stability index, judging whether the stability of each interval meets the normal operating conditions;

[0048] If the stability index value is small and has no obvious changes, it is determined that there is no fault;

[0049] If the stability index fluctuates within the threshold range and the fluctuation amplitude is large, it is determined to be a ground fault;

[0050] If the stability index value fluctuates greatly and exceeds the preset threshold, it is determined to be a cross-line fault;

[0051] The stability indexes of all intervals are traversed, the fault types of each interval are comprehensively analyzed, and finally the specific fault position of the AC side of the AC / DC cross-line fault is located.

[0052] A system for locating an AC-side fault of an AC-DC cross-line fault, the system comprising:

[0053] A real-time monitoring unit obtains the three-phase AC current value in real time and monitors whether the DC bipolar network and the three-phase AC network have fluctuations;

[0054] The information decomposition unit decomposes the three-phase AC current value if a change occurs, and obtains the low-frequency time domain curve after decomposition;

[0055] The stability analysis unit establishes a fault location analysis model, performs stability analysis on the low-frequency time domain curve, and calculates the stability index;

[0056] The fault judgment unit divides the low-frequency time domain curve into intervals, traverses the interval division results based on the stability index, and determines the type of fault occurring in each interval.

[0057] A device for locating an AC-side fault of an AC-DC cross-line fault, the device being used to implement a method for locating an AC-side fault of an AC-DC cross-line fault.

[0058] The technical solution of the present invention can achieve the following technical effects:

[0059] This solves the problem that when an AC / DC cross-line fault occurs in the power grid system, there is no need to check the entire line to determine the fault end. Only the AC current information needs to be collected to determine the fault location and fault type.

[0060] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the embodiments of the present invention 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 recorded in the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0062] Figure 1 It is a flow chart of the method for locating the fault on the AC side of an AC / DC cross-line fault;

[0063] Figure 2 It is the equivalent model diagram of AC and DC power grid;

[0064] Figure 3 It is the timing diagram of the AC current of the intact phase in the three-phase AC network;

[0065] Figure 4 is a timing diagram of the AC current of the ground fault phase in the three-phase AC network;

[0066] Figure 5 is a timing diagram of AC current in a cross-line fault phase in a three-phase AC network;

[0067] Figure 6 The figure is a specific flow chart of the fault location method on the AC side of an AC / DC cross-line fault. DETAILED DESCRIPTION

[0068] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0069] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0070] Embodiment 1;

[0071] like Figure 1 As shown, the present application provides a method for locating an AC-side fault of an AC-DC cross-line fault, the method comprising:

[0072] S10: acquiring the three-phase AC current value in real time and monitoring whether the DC bipolar network and the three-phase AC network have fluctuations;

[0073] S20: if there is a change, decompose the three-phase AC current value to obtain a low-frequency time domain curve after decomposition;

[0074] S30: Establish a fault location analysis model, perform a stability analysis on the low-frequency time domain curve, and calculate a stability index;

[0075] S40: Divide the low-frequency time domain curve into intervals, traverse the interval division results based on the stability index, and determine the type of fault occurring in each interval.

[0076] Specifically, current sensors (such as current transformers) installed at various nodes of the power grid are used to collect three-phase AC current signals in real time. The signal collection frequency is set to 1000 times per second or higher to ensure that subtle changes in current can be captured, and the state of the DC bipolar network, including its current and voltage values, is monitored in real time. Special attention is paid to the fluctuation of current, and the mutual influence between AC and DC power grids and possible cross-line faults are analyzed; the collected three-phase AC current signals are decomposed and processed at multiple levels using wavelet transform. Wavelet transform can decompose the current signal into components of different frequency bands, with special attention paid to low-frequency signals (such as layer a5). The low-frequency time domain signal obtained by wavelet transform contains information about the long-term trend of current changes and system stability; the fault location analysis model is constructed based on the fluctuation characteristics of the current signal, and the stability analysis method is used to evaluate the stability of the low-frequency time domain curve. The stability analysis determines whether the signal is stable by calculating the standard deviation, variance and other indicators of the signal. The calculation of the stability index is based on the low-frequency signal after wavelet transform. If the fluctuation amplitude is large and continuous, it may be a sign of system failure; according to the indicators obtained by the stability analysis, multiple intervals are set for Different fluctuation characteristics of the current signal are distinguished. Each interval represents a specific range of signal change and is divided according to the fluctuation situation. The interval division standard can be set as the change rate based on the smoothness index to ensure that each interval can accurately reflect the change process of the current signal; for each interval after division, the fault type is judged according to the smoothness index and the fluctuation mode. If the fluctuation in a certain interval exceeds the set threshold range, it is determined that there is a fault in the interval, and its type is further determined. Cross-line faults are usually manifested as a sharp change in the current fluctuation amplitude, while grounding faults usually present a relatively gentle fluctuation. The fault type of each interval is classified according to the fluctuation amplitude and change frequency.

[0077] The technical solution of the present invention solves the problem that when an AC / DC cross-line fault occurs in a power grid system, it is not necessary to check the entire line to determine the fault end, and only the AC current information needs to be collected to determine the fault location and fault type.

[0078] Furthermore, the fault location analysis model is established, including:

[0079] Analyze the topology, operating status and current characteristics of DC bipolar networks and three-phase AC networks and establish initial spatial models;

[0080] Based on the initial spatial model, random fault simulation is performed to generate current fluctuation characteristics under different conditions;

[0081] Analyze the current fluctuation characteristics and sort out the trigger conditions, and generate a fault location analysis model based on the trigger conditions training;

[0082] Perform random fault simulation again, continuously optimize model parameters and fault diagnosis algorithm based on fault simulation results, and iteratively update the fault location analysis model.

[0083] As a preferred embodiment of the above, a preliminary model of the power grid is constructed by analyzing the topological structures of the DC bipolar network and the three-phase AC network. The topological structure analysis includes determining the connection relationship between various power grid components (such as generators, transformers, feeders, etc.) and understanding the interaction between the DC system and the AC system; evaluating the operating status of the power grid, including factors such as the current load status of the power grid, the direction of current flow, and frequency fluctuations, so as to provide a basis for subsequent fault simulation. Based on the topological structure and operating status of the power grid, the current characteristics are analyzed, including the current size and direction of each power grid node and the current fluctuation characteristics in different time periods, so as to identify the normal fluctuation range of the current; the above analysis results are used to establish an initial spatial model, which includes various key parameters of the power grid and provides basic data for simulating faults; according to the established initial spatial model, random fault simulation is performed, and the simulation process includes: randomly selecting the fault type, randomly selecting the fault location, and randomly setting the transition resistance. By simulating the current fluctuations generated, current fluctuation characteristics are generated. These characteristics include changes in current amplitude, fluctuation period, and frequency distribution; the generated current fluctuation characteristics are analyzed to identify key trigger conditions, which include critical values ​​of current amplitude changes, frequency fluctuations, and current fluctuations. The fluctuation range and time characteristics are used to provide training data for fault diagnosis according to these trigger conditions, and the basic data for generating the fault location analysis model is sorted out. Based on the collected current fluctuation characteristics and trigger conditions, the fault location analysis model is generated through machine learning, neural network or other intelligent algorithm training. By learning the relationship between current fluctuation characteristics and fault types, unknown faults can be accurately identified. During the training process, the model continuously optimizes its fault diagnosis algorithm to improve the accuracy and response speed of fault location. Different types of fault fluctuation characteristics and trigger conditions are used as input data for training, and the algorithm parameters (such as weights, biases, etc.) are continuously adjusted to adapt to different fault scenarios. By performing random fault simulation again, different fault types, locations and transition resistances are used to generate new current fluctuation characteristics. According to the simulation results and the performance of the analysis model, the model parameters are iteratively updated, and the fault diagnosis algorithm is continuously optimized to ensure its accuracy and reliability in practical applications. After completing the training and optimization of the model, the fault is located. By inputting real-time current data into the fault location analysis model, the system can quickly identify the fault type and fault location, provide specific area information where the fault occurs, and help with the emergency response of the power grid.

[0084] Further, random fault simulation is performed, including:

[0085] Randomly select fault locations in DC bipolar networks and three-phase AC networks;

[0086] Randomly determine any fault type of a cross-line fault or a ground fault, and define different current fluctuation characteristics according to the fault type;

[0087] Set different transition resistance parameters for each fault type to simulate different fault severity and generate multiple current fluctuation modes;

[0088] Based on the current fluctuation characteristics and current fluctuation patterns, the fault location analysis model is preliminarily trained to locate the fault and classify the types of subsequent fault scenarios.

[0089] As a preferred embodiment of the above embodiment, according to the topological structure of the AC / DC hybrid power grid, different fault locations of a DC bipolar network (including a positive pole and a negative pole) and a three-phase AC power grid are simulated. In each simulation, a fault point is randomly selected. The fault location can be simulated by randomly generated coordinates or by selecting a typical fault location (for example, the middle of the line, a branch, etc.) from actual power grid data. The fault point can be set to different parts of the power grid, such as the load end of the DC system, the transmission line of the AC system, the AC / DC connection point, etc.; for each simulated fault location, any type of cross-line fault (such as DC current entering the AC network) and ground fault (such as current entering the ground through a grounding point) is randomly selected for simulation. According to the selected fault type, the corresponding current fluctuation characteristics are defined. For example, for a cross-line fault, the current fluctuation is more violent and the fluctuation amplitude is large; while for a ground fault, the current fluctuation is smaller and the frequency changes slowly; different transition resistance parameters are set for each fault type, and the transition resistance determines the fault current. The size and change rate of the current, for example, for a ground fault, a smaller transition resistance is set to simulate the current flowing quickly to the ground; for a cross-line fault, a larger transition resistance is set to simulate the impedance of the current when it is transmitted between different power grids. In each simulation scenario, different current fluctuation patterns are generated according to different transition resistance parameters. By adjusting the size of the transition resistance, faults of different severity are simulated, thereby ensuring the generation of diverse fault data; preliminary training is carried out based on the current fluctuation characteristics and current fluctuation patterns generated by the simulation, and the current signal (including its fluctuation pattern) is used as input data to train the fault location analysis model to identify the type and location of the fault. During the training process, the model learns by identifying the current fluctuation characteristics of each fault scenario (such as amplitude changes, frequency distribution, etc.), and gradually optimizes the fault classification and location capabilities. During the training process, the model will be adjusted for different fault types and transition resistance parameters to ensure its adaptability and accuracy to future fault scenarios.

[0090] To go further, the three-phase AC current value is decomposed, including:

[0091] Wavelet transform is used to perform multi-layer decomposition on the three-phase AC current value and extract the decomposed signals of different frequency bands;

[0092] Extract the main fluctuation information and fault characteristics from the decomposed low-frequency time domain curve;

[0093] Analyze the high-frequency time domain curve, evaluate the detailed changes of the current signal, and assist in distinguishing high-frequency noise from fault signals in the three-phase AC network;

[0094] Based on the decomposition results of the low-frequency time domain curve and the high-frequency time domain curve, the stability analysis is carried out in combination with the time characteristics of the three-phase AC current value.

[0095] As a preferred embodiment of the above, the three-phase current signal (A, B, C phases respectively) of the AC power grid is collected in real time through current transformers (CT) or Hall sensors and other devices, and the sampling frequency of the signal acquisition system is set to at least 1000 Hz to ensure high-precision capture of instantaneous changes in the current signal; the collected three-phase AC current value is decomposed at multiple levels by wavelet transform to extract signals of different frequency bands. The wavelet transform performs multi-scale decomposition of the signal based on the mother wavelet function (such as Morlet wavelet, Daubechies wavelet, etc.), and the decomposition result is a series of signals of different frequency bands, each of which is a wavelet of a plurality of frequencies. One layer represents the performance of the current signal at different scales. For example, the current signal is decomposed using wavelet transform. The extracted low-frequency signal (such as layer a5) reflects the long-term trend of the signal and the stability of the system, while the high-frequency signal (such as layers d1 and d2) captures short-term changes and details. The main fluctuation information and fault characteristics are extracted from the low-frequency time domain curve. The main fluctuation information of the signal is extracted from the decomposed low-frequency time domain curve. These fluctuation information include: amplitude changes that reflect the amplitude changes of the current signal. Large amplitude changes may indicate certain abnormalities or faults in the system; evaluate the fluctuation period of the current signal The characteristics of the current signal are analyzed to help determine whether the system is running stably; the frequency distribution of the low-frequency signal is analyzed to further extract the timing characteristics of the signal and match it with the characteristics of the fault signal; the high-frequency time domain curve (such as the signals of the d1 and d2 layers) is analyzed to evaluate the detailed changes of the current signal. The high-frequency band usually contains instantaneous changes in the current signal, such as rapid fluctuations, noise or small disturbances of the system. Analyzing these high-frequency changes can help determine whether it is caused by high-frequency noise (such as electromagnetic interference during equipment operation) or actual signal changes caused by faults. The threshold judgment method is used to determine whether the low-frequency signal is caused by the actual signal. The fluctuation amplitude and frequency change are used to determine whether it is a fault feature; the decomposition results of the low-frequency and high-frequency time domain curves are combined, and the time characteristics of the three-phase AC current value (i.e., the timing changes of the signal) are considered to perform a stationarity analysis. The stationarity analysis evaluates the stability of the signal by calculating the statistical characteristics of the signal, such as the standard deviation and mean. If the signal fluctuates greatly and lasts for a long time, it may indicate a system fault. The stability of the signal is determined based on the time domain characteristics. If the signal continues to be unstable or there is a sudden change, the fault type is further determined, such as a cross-line fault or a ground fault; the fault type is determined based on the stationarity analysis and the change in fluctuation information. For example, if the fluctuation amplitude of the low-frequency signal is large and accompanied by a sudden change in the high-frequency signal, it is determined to be a cross-line fault; if the low-frequency signal fluctuates less and the high-frequency signal changes more smoothly, it is determined to be a ground fault. According to the analysis of the fault type and current signal, the fault is located to determine the area and specific location of the fault.

[0096] Furthermore, the main fluctuation information is extracted from the decomposed low-frequency time domain curve, including:

[0097] According to the wavelet transform, the low-frequency time domain curve is multi-layered decomposed, and the low-level wavelet frequency band time domain image after the low-frequency time domain curve is decomposed is obtained. The time domain signal is defined as f(t). The specific wavelet transform is:

[0098]

[0099] Where W f (a,b) represents the transform coefficient of the signal f(t) at scale a and position b, is the mother wavelet function, which is transformed by the scale factor a and the translation factor b They are collectively called wavelets;

[0100] For discrete time series:

[0101]

[0102] The overall formula represents the function of the mother wavelet in the discrete time series after scale transformation j and translation transformation k. represents the sub-wavelet function after scale and translation transformation, 2 -j / 2 represents the scale factor, -k represents the translation factor;

[0103] The main fluctuation information of the low-frequency time domain curve is obtained through wavelet transform, including the amplitude change, fluctuation period and frequency distribution of the signal;

[0104] The fault characteristics in the current signal are determined by combining the main fluctuation information of the low-frequency time domain curve with the analysis results of the high-frequency time domain curve.

[0105] As a preferred embodiment of the above, a current sensor (such as a current transformer CT, a Hall sensor, etc.) is used to obtain three-phase AC current values ​​(phases A, B, and C) in real time; the collected three-phase AC current signal is subjected to wavelet transform, and the signal is decomposed at multiple levels using appropriate wavelet basis functions (such as Morlet wavelet, Daubechies wavelet, etc.). By decomposing, the low-frequency part (for example, a5 layer) and high-frequency part (for example, d1, d2 layers) of the signal are extracted, and the time domain signal is transformed using wavelet basis functions. For discrete time series, discrete wavelet transform (DWT) is used for wavelet transform. Through this transformation, the time domain signal is decomposed into frequency bands of different scales, and the low-frequency and high-frequency components can be clearly extracted; the main fluctuation information is extracted from the decomposed low-frequency band time domain curve, including: amplitude change, fluctuation period, frequency distribution; the decomposed high-frequency band time domain curve (such as d1, d2 and other layers) is analyzed. The high-frequency band time domain signal usually contains instantaneous changes and noise. Analyzing its fluctuation characteristics helps to distinguish between fault signals and normal high-frequency noise; by calculating the fluctuation amplitude and frequency change of the high-frequency signal, it is determined whether the change of the current signal is caused by a fault or by external electromagnetic interference or equipment noise. The fault characteristics in the current signal are further determined by combining the main fluctuation information of the low-frequency time domain curve and the analysis results of the high-frequency time domain curve. If the fluctuation modes of both the low-frequency and high-frequency bands show abnormal changes (such as a sharp increase in amplitude, periodic changes, etc.), there may be a cross-line fault. If the low-frequency fluctuation is small and the high-frequency signal shows continuous instability, it may be a ground fault. By comparing these fluctuation characteristics with the known fault modes, the fault type is finally confirmed. According to the fluctuation information extracted from the low-frequency and high-frequency time domain curves, combined with the results of wavelet transform, the fault type is judged and located. Cross-line faults are usually manifested as large low-frequency fluctuations and obvious high-frequency mutations. The ground fault has small low-frequency fluctuations, and the high-frequency fluctuations are relatively stable or show a continuous growth trend. Combining all analysis results, the fault location and type are accurately located.

[0106] Further, the calculation of stationarity indicators includes:

[0107] Calculate the average value of the low-frequency time domain curve and calculate the stability index based on the main fluctuation information:

[0108]

[0109] Among them, S(I i ) represents the stability index, x i represents the i-th sample, represents the mean value, and N represents the total number of data points;

[0110] According to the smoothness index, the fault type is judged by interval division, and the smoothness index of each interval is compared with the preset fault threshold to determine whether the fault condition is met.

[0111] As a preferred embodiment of the above embodiment, the collected three-phase AC current signal is decomposed at multiple levels through wavelet transform to extract low-frequency signals, and the three-phase current signal is decomposed using a wavelet transform algorithm (such as Daubechies wavelet, Morlet wavelet, etc.), and the low-frequency time domain curve (such as a5 layer signal) is extracted, and the average value of the low-frequency time domain curve is calculated. According to the main fluctuation information of the low-frequency time domain curve, the smoothness index is calculated; according to the calculated smoothness index, multiple intervals are set, and the smoothness value of each interval is compared with a preset fault threshold. When the smoothness index is lower than the preset normal fluctuation range, the interval is determined to be in a normal operating state; when When the smoothness index exceeds the set threshold, it is determined to be a fault interval; when dividing the intervals, determine whether the interval conforms to known fault modes such as ground faults and cross-line faults based on the fluctuation mode and fluctuation amplitude of each interval; determine the fault type of each interval based on the interval division results and fault standards. For example, if the smoothness index shows a large fluctuation and exceeds the normal range, and the fluctuation amplitude of the high-frequency signal increases, it can be determined as a cross-line fault; if the smoothness index changes in a small amplitude, and the high-frequency band fluctuations are relatively stable, it is determined to be a ground fault. By traversing all intervals, the smoothness index and fluctuation characteristics are comprehensively analyzed, and finally the fault type of each interval is confirmed, and the specific location of the fault is located.

[0112] Furthermore, the low-frequency time domain curve is divided into intervals, including:

[0113] According to the stability index of the low-frequency time domain curve, a preset threshold range is determined, and multiple intervals are set to divide different fluctuation characteristics;

[0114] According to the change of the stability index, the low-frequency time domain curve is divided into intervals according to the threshold range, and the fluctuation pattern corresponding to each interval is generated;

[0115] Analyze each divided interval to determine whether the fluctuation characteristics of each interval meet the preset fault standards;

[0116] If the fluctuation within the interval exceeds the set threshold, the type of fault occurring in each interval is determined based on the fluctuation pattern within the interval.

[0117] As a preferred embodiment of the above embodiment, the collected three-phase current signal is decomposed by wavelet transform to extract the low-frequency time domain curve (such as a5 layer signal). These low-frequency signals mainly reflect the long-term trend of the current signal and the stability of the system; the average value of the low-frequency time domain curve is calculated, and the smoothness index is calculated according to the fluctuation information of the low-frequency time domain curve. The index quantifies the volatility of the signal and evaluates its stability. The specific calculation method is to analyze whether the signal is stable based on the fluctuation amplitude of the current signal; according to the calculated smoothness index, the threshold range is set, and the low-frequency time domain curve is divided into intervals. The purpose of interval division is to identify different fluctuation feature intervals according to the changes in the smoothness index. The fluctuation information in each interval will be analyzed separately, and a corresponding fluctuation pattern will be generated for each interval. These fluctuation patterns help determine whether the signal meets the normal operating state or a fault occurs. The fluctuation characteristics within the interval are compared with the preset fault standard. If the fluctuation in the interval exceeds the set threshold, the interval is judged to be a fault interval. The fault type is determined by analyzing the fluctuation pattern of each interval. If the fluctuation characteristics within the interval match the characteristics of a ground fault or a cross-line fault, the fault type of the interval is confirmed and the fault is located. All intervals are traversed, and the smoothness indicators and fluctuation patterns of each interval are comprehensively analyzed to confirm the fault type and the specific location where it occurs.

[0118] To further explain, the fault type in each interval is determined, including:

[0119] According to the stability index, determine whether the stability of each interval meets the normal operating conditions;

[0120] If the stability index value is small and has no obvious changes, it is determined that there is no fault;

[0121] If the stability index fluctuates within the threshold range and the fluctuation amplitude is large, it is determined to be a ground fault;

[0122] If the stability index value fluctuates greatly and exceeds the preset threshold, it is determined to be a cross-line fault;

[0123] Traverse the stability indicators of all intervals, comprehensively analyze the fault type of each interval, and finally locate the specific fault location on the AC side of the AC / DC cross-line fault.

[0124] As a preferred embodiment of the above, a DC bipolar network P-phase-AC network A-phase cross-line fault scenario is selected, and the fault point on the DC bipolar network line is F 1 , the fault point on the AC network line is F 2 , the transition resistance is R f In this fault scenario, the current values ​​of the three-phase AC M side and AC N side in the AC network A, B and C are randomly collected and recorded as I i ; Then the AC current value Ii Perform wavelet decomposition, take the low-frequency wavelet band time domain curve after decomposition, and record this curve as L(I i ), and for the curve L(I i ) for stationarity analysis and is recorded as S(I i ); Finally, S(I i ) is divided into intervals. If S(I i )∈[0,0.1], it is recorded that this phase has no fault; if S(I i )∈(0.1,0.2], it is recorded as a phase-to-ground fault; if S(I i )>0.2, it is recorded as a cross-line fault in this phase.

[0125] Embodiment 2:

[0126] Based on the same inventive concept as the method for locating the AC side fault of an AC / DC cross-line fault in the aforementioned embodiment, the present invention further provides a system for locating the AC side fault of an AC / DC cross-line fault, the system comprising:

[0127] A real-time monitoring unit obtains the three-phase AC current value in real time and monitors whether the DC bipolar network and the three-phase AC network have fluctuations;

[0128] The information decomposition unit decomposes the three-phase AC current value if a change occurs, and obtains the low-frequency time domain curve after decomposition;

[0129] The stability analysis unit establishes a fault location analysis model, performs stability analysis on the low-frequency time domain curve, and calculates the stability index;

[0130] The fault judgment unit divides the low-frequency time domain curve into intervals, traverses the interval division results based on the stability index, and determines the type of fault occurring in each interval.

[0131] The above adjustment system in the present invention can effectively implement a method for locating the AC side fault of an AC / DC cross-line fault, and the technical effects that can be achieved are as described in the above embodiments, which will not be repeated here.

[0132] Embodiment three;

[0133] Based on the same inventive concept as the method for locating an AC side fault of an AC / DC cross-line fault in the aforementioned embodiment, the present invention also provides an AC side fault locating device for an AC / DC cross-line fault, and the device is used to implement the method for locating an AC side fault of an AC / DC cross-line fault.

[0134] The device in the present invention can effectively implement a method for locating the AC side fault of an AC / DC cross-line fault, and the technical effects that can be achieved are as described in the above embodiments and will not be repeated here.

[0135] Although the present application has been described in conjunction with specific features and embodiments thereof, it is obvious that various modifications and combinations may be made thereto without departing from the spirit and scope of the present application. Accordingly, this specification and the accompanying drawings are merely exemplary illustrations of the present application as defined therein, and are deemed to have covered any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application is intended to include these modifications and variations.

Claims

1. A method for locating the AC side fault of an AC / DC cross-line fault, characterized in that: The method comprises: Obtain three-phase AC current values ​​in real time and monitor whether there are fluctuations in the DC bipolar network and the three-phase AC network; If there is a change, the three-phase AC current value is decomposed to obtain a low-frequency time domain curve after decomposition; Establishing a fault location analysis model, performing a stability analysis on the low-frequency time domain curve, and calculating the stability index; The low-frequency time domain curve is divided into intervals, and the interval division results are traversed based on the smoothness index to determine the type of fault occurring in each interval.

2. The method for locating the AC side fault of an AC / DC cross-line fault according to claim 1, characterized in that: Establish a fault location analysis model, including: Analyzing the topological structure, operating status and current characteristics of the DC bipolar network and the three-phase AC network and establishing an initial spatial model; Based on the initial spatial model, random fault simulation is performed to generate current fluctuation characteristics under different conditions; Analyze the current fluctuation characteristics and sort out the trigger conditions, and generate a fault location analysis model based on the training of the trigger conditions; The random fault simulation is performed again, model parameters and fault diagnosis algorithms are continuously optimized according to the fault simulation results, and the fault location analysis model is iteratively updated.

3. The method for locating the AC side fault of an AC / DC cross-line fault according to claim 2, characterized in that: Perform random fault simulations, including: Randomly select fault locations in DC bipolar networks and three-phase AC networks; Randomly determine any fault type of a cross-line fault and a ground fault, and define different current fluctuation characteristics according to the fault type; Setting different transition resistance parameters for each of the fault types, simulating different fault severities, and generating multiple current fluctuation modes; According to the current fluctuation characteristics and the current fluctuation pattern, the fault location analysis model is preliminarily trained to perform fault location and type classification for subsequent fault scenarios.

4. The method for locating the AC side fault of an AC / DC cross-line fault according to claim 1, characterized in that: Decomposing the three-phase alternating current value includes: Using wavelet transform to perform multi-layer decomposition on the three-phase AC current value, and extracting the decomposed signals of different frequency bands; Extracting main fluctuation information and fault characteristics from the low-frequency time domain curve obtained by decomposition; Analyze the high-frequency time domain curve, evaluate the detailed changes of the current signal, and assist in distinguishing high-frequency noise from fault signals in the three-phase AC network; The stability analysis is performed based on the decomposition results of the low-frequency time domain curve and the high-frequency time domain curve in combination with the time characteristics of the three-phase alternating current value.

5. The method for locating the AC side fault of an AC / DC cross-line fault according to claim 4, characterized in that: Extracting the main fluctuation information from the decomposed low-frequency time domain curve includes: The low-frequency time domain curve is multi-layered decomposed according to wavelet transform, and the low-level wavelet frequency band time domain image after the decomposition of the low-frequency time domain curve is obtained, and the time domain signal is defined as f(t). The wavelet transform is specifically as follows: Where W f (a,b) represents the transform coefficient of the signal f(t) at scale a and position b, is the mother wavelet function, which is transformed by the scale factor a and the translation factor b They are collectively called wavelets; For discrete time series: The overall formula represents the function of the mother wavelet in the discrete time series after scale transformation j and translation transformation k. represents the sub-wavelet function after scale and translation transformation, 2 -j / 2 represents the scale factor, -k represents the translation factor; Obtaining the main fluctuation information of the low-frequency time domain curve by wavelet transform, including the amplitude change, fluctuation period and frequency distribution of the signal; The main fluctuation information of the low-frequency time domain curve is combined with the analysis result of the high-frequency time domain curve to determine the fault characteristics in the current signal.

6. The method for locating the AC side fault of an AC / DC cross-line fault according to claim 1, characterized in that: Calculating the stability index includes: Calculate the average value of the low-frequency time domain curve and calculate the stability index based on the main fluctuation information: Among them, S(I i ) represents the stability index, x i represents the i-th sample, represents the average value, and N represents the total number of data points; According to the smoothness index, the fault type is judged by interval division, and the smoothness index of each interval is compared with a preset fault threshold to determine whether the fault condition is met.

7. The method for locating the AC side fault of an AC / DC cross-line fault according to claim 1, characterized in that: The low-frequency time domain curve is divided into intervals, including: According to the stability index of the low-frequency time domain curve, a preset threshold range is determined, and a plurality of intervals are set for dividing different fluctuation characteristics; According to the change of the stability index, the low-frequency time domain curve is divided into intervals according to the threshold range, and a fluctuation pattern corresponding to each interval is generated; Analyze each divided interval to determine whether the fluctuation characteristics of each interval meet the preset fault standard; If the fluctuation within the interval exceeds the set threshold, the fault type occurring in each interval is determined based on the fluctuation pattern within the interval.

8. The method for locating the AC side fault of an AC / DC cross-line fault according to claim 7, characterized in that: Determine the type of fault that occurs in each interval, including: According to the stability index, judging whether the stability of each interval meets the normal operating conditions; If the stability index value is small and has no obvious changes, it is determined that there is no fault; If the stability index fluctuates within the threshold range and the fluctuation amplitude is large, it is determined to be a ground fault; If the stability index value fluctuates greatly and exceeds the preset threshold, it is determined to be a cross-line fault; The stability indexes of all intervals are traversed, the fault types of each interval are comprehensively analyzed, and finally the specific fault position of the AC side of the AC / DC cross-line fault is located.

9. A system for locating the AC side fault of an AC / DC cross-line fault, characterized in that: The system comprises: A real-time monitoring unit obtains the three-phase AC current value in real time and monitors whether the DC bipolar network and the three-phase AC network have fluctuations; The information decomposition unit decomposes the three-phase AC current value if a change occurs, and obtains the low-frequency time domain curve after decomposition; The stability analysis unit establishes a fault location analysis model, performs stability analysis on the low-frequency time domain curve, and calculates the stability index; The fault judgment unit divides the low-frequency time domain curve into intervals, traverses the interval division results based on the stability index, and judges the type of fault occurring in each interval.

10. A device for locating the AC side fault of an AC / DC cross-line fault, characterized in that: The device is used to implement the method according to any one of claims 1 to 8.

Images