Artificial intelligence analysis method and system applied to traveling wave fault location

Abstract

The invention discloses an artificial intelligence analysis method and system applied to traveling wave fault positioning, and relates to the technical field of power fault positioning. Traveling wave data and standing wave data are collected through a current sensor and a voltage sensor and are preprocessed; a convolutional neural network and long and short term memory network mixed model is combined with a space-time attention mechanism, and space-time features and association complementarity of two data modes are deeply mined; based on fusion features, fault types are divided into typical faults and complex faults, and the typical faults and the complex faults are identified and positioned respectively; typical faults are positioned through traveling wave time difference, and standing wave verification is carried out. Carrying out complex fault analysis and multi-wave head and standing wave reflection characteristic comprehensive positioning; when other fault points exist near the fault point, analyzing the time sequence and space relevance among the multiple fault points; and in combination with fault types, positioning and relevance information, operation and maintenance personnel are assisted to make decisions and diagnoses.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of power fault positioning, in particular to an artificial intelligence analysis method and system applied to traveling wave fault positioning. BACKGROUND

[0002] A power transmission line is an important component of a power system, and its safe and stable operation is crucial for normal power supply. However, various types of faults are inevitable during the long-term operation of the power transmission line.

[0003] When a fault occurs, a traveling wave is generated at the fault point, which propagates along the line and is reflected and refracted at uneven points on the line to form a standing wave. Traditional fault positioning techniques mainly focus on the capture and analysis of the initial traveling wave and consider establishing a correlation model with the generated standing wave and comprehensively analyzing the correlation and complementarity. Traditional techniques often only analyze and calculate the fault type and positioning uniformly, and do not make case-by-case judgments according to different types. At the same time, there is a lack of in-depth analysis of the traveling wave grounding characteristics and the zero sequence component in the standing wave, which leads to insufficient identification and inaccurate results. When there are other fault points near the fault location, the existing technology does not consider the correlation between similar faults, which affects the efficiency of fault diagnosis and processing. SUMMARY

[0004] The purpose of the present application is to provide an artificial intelligence analysis method and system applied to traveling wave fault positioning to solve the problems in the prior art.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical solutions: In a first aspect, the present application provides an artificial intelligence analysis method applied to traveling wave fault positioning, comprising: Collecting traveling wave data and standing wave data through current sensors and voltage sensors, and preprocessing; Using a hybrid model combining a convolutional neural network and a long short-term memory network to extract spatiotemporal features from the preprocessed traveling wave and standing wave data, incorporating a spatiotemporal attention mechanism, mining the correlation and complementarity between the two modal features, and realizing deep fusion of information; Based on the fused features, using the hybrid model for fault type identification and fault positioning; combining the traveling wave grounding characteristics and the zero sequence component in the standing wave to identify the fault type of a typical fault; and fusing and analyzing the information dimension provided by the feature combination of the traveling wave and the standing wave to obtain the fault type of a complex fault; In terms of fault positioning, for a typical fault, the time difference between the arrival of the initial traveling wave head at the two end measurement points is focused on, the fault distance is calculated in combination with the line parameters, and the standing wave information is used as auxiliary verification; for a complex fault, the time interval and energy distribution of multiple wave heads are analyzed, the fault point reflection characteristics reflected in the standing wave information are relied on, and the fault location is comprehensively judged; When multiple fault points are detected near the fault location, a correlation analysis is performed using a hybrid model to evaluate whether there is a time sequence and a spatial propagation relationship between the fault signals, and fault correlation is obtained; Based on the fault type, fault location and fault correlation, the operation and maintenance personnel are assisted in decision-making and diagnosis.

[0006] In combination with the first aspect, in a first implementation manner of the first aspect of the application, the acquisition of the traveling wave data and the standing wave data by the current sensor and the voltage sensor, and the preprocessing, include: Current sensors and voltage sensors are deployed at key nodes of the power transmission line to measure the traveling wave transient current and voltage signals at the key nodes, and to collect the voltage and current signals under steady state and quasi-steady state after the fault; The collected raw data is preprocessed, data alignment and synchronization verification are performed, the GPS timestamp is used to adjust the time alignment of the data collected by different sensor nodes to make the data synchronized in the time dimension, digital filtering technology is used for data denoising and filtering to remove high-frequency noise and power frequency interference, wavelet transform method is used to analyze the traveling wave signal to extract the instantaneous amplitude and highlight the key features and instantaneous changes related to the fault in the signal, and data loss and abnormal points are processed and corrected by using the mean filling method.

[0007] In combination with the first aspect, in a second implementation manner of the first aspect of the application, the hybrid model combining the convolutional neural network and the long short-term memory network is used to extract the spatiotemporal features from the preprocessed traveling wave and standing wave data, to integrate the spatiotemporal attention mechanism, to mine the correlation and complementarity between the two modal features, and to realize the deep fusion of information, including: The hybrid model includes a convolutional neural network CNN branch and a long short-term memory network LSTM branch, the preprocessed data is taken as the input of the hybrid model, the CNN branch captures the structured information of the signal in the local area, the convolution layer operates on the time series of the signal through the sliding filter to automatically learn and extract the local spatial features in the signal, and the pooling layer is used to reduce the feature dimension and enhance the translation invariance of the hybrid model; the LSTM branch captures the evolution features of the signal in the time dimension, processes the time sequence dynamics in the sequence data, and learns and remembers the dependency relationship and dynamic pattern in the long time sequence through the gating mechanism, including the input gate, the forgetting gate and the output gate; After extracting the spatial features and the time sequence features in two branches respectively, the outputs of the two branches are fused and input into a space-time attention mechanism module to dynamically calculate and distribute weights according to the importance of the features for fault type identification and positioning; the importance of the features in the time axis and the importance of the spatial dimensions are evaluated, so that the hybrid model can automatically focus on the most relevant space-time region, suppress noise and redundant information, and effectively mine the relevance and complementarity between the traveling wave signals and the standing wave signals.

[0008] In combination with the first aspect, in a third implementation manner of the first aspect of the present application, the identification of the line wave grounding feature and the zero sequence component in the standing wave is combined to obtain the fault type of the typical fault, including: The features fused through the space-time attention mechanism are used to identify and locate the fault type by combining the hybrid model; in the fault type identification link, the hybrid model focuses on analyzing the line wave grounding transient characteristics and the zero sequence component information in the standing wave reflected in the fused features to distinguish the typical fault types; the typical fault types include single-phase grounding, two-phase short-circuit grounding, two-phase short-circuit and three-phase short-circuit; In identifying the single-phase grounding fault, the initial transient signal of the line wave generated by the single-phase grounding fault is captured, which is manifested as that the fault phase voltage is lower than a preset single-phase grounding fault phase voltage threshold, the non-fault phase voltage is higher than a preset single-phase grounding non-fault phase voltage threshold, and the fault phase generates an initial line wave current with an amplitude exceeding a preset single-phase threshold; the hybrid model can identify this initial disturbance feature of single-phase asymmetry, including the difference in the line wave amplitude and polarity between the fault phase and the non-fault phase; during the steady state or quasi-steady state after the fault, the single-phase grounding fault generates a zero sequence current and a zero sequence voltage higher than a preset single-phase grounding zero sequence threshold, the hybrid model analyzes the frequency spectrum or time domain information of the standing wave part in the fused features to detect the existence and characteristics of the zero sequence component; since the zero sequence component has a specific amplitude and mode when single-phase grounding; when the initial line wave single-phase asymmetry disturbance is detected, and it is confirmed that the zero sequence component in the standing wave exists in the single-phase grounding mode, the single-phase grounding fault is determined; In identifying the two-phase short-circuit-to-ground fault, the initial transient signal of the traveling wave is captured when two phases are simultaneously short-circuited and at least one phase is grounded. The traveling wave signal is characterized in that the voltages of the two fault phases are lower than the preset two-phase short-circuit-to-ground fault phase voltage threshold, and the voltage of the non-fault phase is higher than or equal to the preset two-phase short-circuit-to-ground non-fault phase voltage threshold. The hybrid model analyzes the initial transient signal of the traveling wave, learns and identifies the specific characteristics of the phase relationship of the two fault phases under this fault, which is characterized in that the initial phase difference is within the preset two-phase phase difference threshold interval. The two phases generate initial traveling wave currents with amplitudes exceeding the preset two-phase threshold, and there is a specific phase relationship. The two-phase short-circuit-to-ground generates zero-sequence current and zero-sequence voltage, but the magnitude and characteristics are different from those of single-phase grounding. The zero-sequence component here is caused by two-phase simultaneous grounding or one-phase grounding and short-circuit of another phase, and has identifiable amplitude and phase characteristics. When the two-phase asymmetry disturbance of the traveling wave signal is detected, and it is confirmed that there is a zero-sequence component in the standing wave that meets the two-phase grounding mode, the two-phase short-circuit-to-ground fault is determined. In identifying the two-phase short-circuit fault, the hybrid model analyzes the initial transient characteristics of the traveling wave signal, which is characterized in that the voltages and currents of the two phases involved are lower than or deviate from the preset two-phase short-circuit fault voltage and current normal range, and the voltages and currents of the non-fault phase are maintained within the normal range. Since there is no grounding, the hybrid model analyzes the fusion characteristics of the standing wave part, and detects the zero-sequence component with an amplitude lower than the preset two-phase short-circuit fault zero-sequence threshold or equal to zero, which is also a key feature for distinguishing. When the hybrid model detects the initial transient characteristics of the two phases in the traveling wave signal, and confirms that the amplitude of the zero-sequence component in the standing wave is lower than the preset two-phase short-circuit fault zero-sequence threshold, the two-phase short-circuit fault is determined. In identifying the three-phase short-circuit fault, the hybrid model analyzes the initial transient characteristics reflected in the fusion characteristics of the three-phase simultaneous short-circuit, and the traveling wave signal is characterized in that the three-phase voltages are simultaneously lower than the respective preset three-phase short-circuit fault voltage threshold, the three-phase currents deviate from the respective preset three-phase short-circuit fault current normal range, and there is a specific phase relationship between the three phases, which is characterized in that the initial change start time difference of the three-phase voltages and currents is less than the preset three-phase time threshold, and the change waveform similarity is higher than the preset three-phase similarity threshold. The hybrid model analyzes the fusion characteristics of the standing wave part, and detects the zero-sequence component with an amplitude lower than the preset three-phase short-circuit fault zero-sequence threshold or equal to zero. When the hybrid model detects the initial transient characteristics of the three phases in the traveling wave signal, and confirms that the amplitude of the zero-sequence component in the standing wave is lower than the preset three-phase short-circuit fault zero-sequence threshold, the three-phase short-circuit fault is determined.

[0009] In combination with the first aspect, in a fourth implementation manner of the first aspect of the present application, the information dimension provided by the combination of the characteristics of the fusion analysis of the traveling wave and the standing wave is used to obtain the fault type of the complex fault, including: The complex fault type includes high-resistance grounding fault and intermittent fault. In identifying the high resistance ground fault, the amplitude of the initial transient signal of the traveling wave in the mixed model analysis fusion feature is lower than the preset high resistance ground fault traveling wave amplitude threshold, the waveform change exceeds the preset high resistance ground fault waveform normal range, and the transient signal is accompanied by noise or oscillation with an amplitude lower than the preset noise threshold and a frequency higher than the preset high frequency threshold; at the same time, in the standing wave analysis, the zero sequence component amplitude is lower than the typical ground fault, the time change rate is lower than the preset zero sequence component change threshold, and the fluctuation exceeds the preset high resistance ground fault zero sequence component fluctuation normal range; the mixed model determines the high resistance ground fault by analyzing the characteristic combination that the traveling wave transient is lower than the preset high resistance ground fault traveling wave amplitude threshold and the standing wave zero sequence component amplitude is low and unstable. In identifying the intermittent fault, the mixed model analyzes the time-varying of the traveling wave signal in the fusion feature, which is manifested as the appearance interval of the traveling wave transient signal of the fault phase voltage and current exceeding the preset minimum interval time, the disappearance interval exceeding the preset maximum disappearance time, the amplitude exceeding the preset intermittent fault amplitude range and the polarity change exceeding the preset polarity rule change range, and the arrival time interval stability being lower than the preset intermittent fault interval stability threshold; at the same time, in the standing wave analysis, the existence time of the zero sequence component is lower than the preset intermittent fault zero sequence component persistence time threshold; the mixed model determines the intermittent fault by capturing the characteristic combination that both the traveling wave transient and the standing wave zero sequence component do not meet the preset continuity or stability and have obvious intermittent characteristics.

[0010] In combination with the first aspect, in a fifth implementation manner of the first aspect of the application, in terms of fault location, for a typical fault, the time difference of the initial traveling wave head arriving at the two end measurement points is focused, the fault distance is calculated in combination with the line parameters, and the standing wave information is used as auxiliary verification, including: After identifying that the fault type is a typical fault, the time difference T d of the initial traveling wave head arriving at the two end measurement points of the line is extracted from the traveling wave data, and the fault distance calculation formula is: ; Wherein, D is the fault distance, represents the distance value from one end of the line to the fault point, c is the wave speed, T d is the time difference of the initial traveling wave head arriving at the two end measurement points of the line. In this process, the steady-state characteristics extracted from the standing wave data are used as auxiliary verification information, which is compared with the expected standing wave reflection characteristics corresponding to the fault distance D calculated based on the traveling wave, to verify the traveling wave positioning result and improve the confidence of the positioning result. The fault distance D is output, the mapping relationship of the fault distance D to the fault location coordinate point is established in combination with the geographical or topological information of the line, and the specific coordinate point of the fault location is obtained through linear interpolation calculation.

[0011] In a sixth implementation form of the first aspect, in the case of a complex fault, the time interval and energy distribution of multiple wave fronts are analyzed based on the reflection characteristics of the fault point reflected in the standing wave information, and the fault location is comprehensively determined, including: In the case of identifying the fault type as a complex fault, the mixed model analyzes the time interval sequence and corresponding energy distribution characteristics of multiple traveling wave wave fronts extracted from the traveling wave data reaching the measurement point, and analyzes the steady-state characteristics extracted from the standing wave data, including the evolution law of the standing wave ratio and the reflection coefficient with frequency and time; Based on the comprehensive analysis of the time interval, energy distribution and standing wave reflection characteristics of multiple wave fronts, a unified multi-dimensional feature vector is formed, and the mixed model establishes a mapping relationship between the fault location and the multi-dimensional feature vector. Using historical fault data, the mixed model learns the feature vectors generated by high-resistance grounding and intermittent faults at different locations, and through training, the mixed model automatically learns and fits the nonlinear relationship between the feature vector and the fault location. The extracted feature vector is input into the trained mixed model, and the specific fault location coordinate value is output through calculation. The coordinate value is the final determined fault positioning coordinate, realizing the location positioning of high-resistance grounding and intermittent faults.

[0012] In a seventh implementation form of the first aspect, in the case of detecting multiple fault points near the fault location, the mixed model is used for correlation analysis to evaluate whether there is a time sequence and a spatial propagation relationship between the fault signals, and the fault correlation is obtained, including: In the case of multiple fault points, the mixed model analyzes the arrival time of each wave front to identify the time relationship between different fault points and determine the propagation order of the fault signals. The mixed model uses a time-space attention mechanism to evaluate the spatial propagation characteristics of the fault signals and determine whether there is a spatial propagation relationship between the fault points, including reflection, refraction and scattering of signals. In terms of time analysis, the mixed model analyzes the time difference of fault wave fronts to infer the occurrence order of fault points and help locate the sequence of multiple fault points. In terms of spatial analysis, the mixed model combines the positions of sensor nodes to evaluate the propagation path of signals, fault distance and fault positioning, and further confirms the spatial correlation of different fault points. Based on the correlation analysis in the time and space dimensions, the mixed model comprehensively determines the mutual influence between multiple fault points and obtains the fault correlation.

[0013] In an eighth implementation form of the first aspect, based on the fault type, fault location and fault correlation, the operation and maintenance personnel are assisted in decision-making and diagnosis, including: Based on the deep analysis of traveling wave signals and standing wave signals based on the mixed model, the fault type is accurately identified, including single-phase grounding, two-phase short-circuit grounding, two-phase short-circuit, three-phase short-circuit, high-resistance grounding and intermittent fault; by feeding back the identified fault type to the operation and maintenance personnel, the nature of the fault is provided, and the corresponding processing measures are determined; Combined with fault type and signal propagation analysis, the mixed model provides fault location information to determine the specific location of the fault; by analyzing the propagation time difference of the traveling wave signal, the zero sequence component in the standing wave and the wave head arrival time, the operation and maintenance personnel are helped to locate the fault point in the power line; When there are multiple fault points, the time and space relationship between the fault points is identified through space-time correlation analysis; the mixed model analyzes the mutual influence between the fault points, judges whether the fault is independent or there is a linkage effect, and assists the operation and maintenance personnel to make decisions according to the fault correlation; Based on the fault type, location and correlation, decision support and maintenance strategies are provided for the operation and maintenance personnel.

[0014] In a second aspect, the application provides an artificial intelligence analysis system applied to traveling wave fault location, comprising: A data acquisition and preprocessing module includes a data acquisition unit and a data preprocessing unit; wherein the data acquisition unit acquires traveling wave data and standing wave data through current sensors and voltage sensors; the data preprocessing unit aligns and synchronizes the data, removes noise and filters, extracts features, and corrects missing and abnormal values; A deep feature extraction and fusion module includes a mixed model structure unit, a feature extraction unit and a feature fusion and attention mechanism unit; wherein the mixed model structure unit constructs a mixed network structure containing a convolutional neural network CNN and a long short-term memory network LSTM, which are respectively used to process spatial features and time sequence features; the feature extraction unit extracts the spatial structure features of the input data using the CNN branch and extracts the time evolution features using the LSTM branch; the feature fusion and attention mechanism unit fuses the spatial and temporal features into spatio-temporal features, and applies the attention mechanism to dynamically emphasize important features and suppress irrelevant information; A fault type identification module includes a typical fault identification unit and a complex fault identification unit; wherein the typical fault identification unit focuses on analyzing the traveling wave grounding transient characteristics and the standing wave zero sequence component information based on the fused features of the space-time attention mechanism, and identifies typical fault types including single-phase grounding, two-phase short-circuit grounding, two-phase short-circuit and three-phase short-circuit; the complex fault identification unit identifies high-resistance grounding faults and intermittent faults by analyzing the feature combinations exhibited by traveling waves and standing waves based on the fused features of the space-time attention mechanism; Fault location module: including typical fault location unit and complex fault location unit; wherein, the typical fault location unit calculates the fault distance based on the traveling wave time difference, verifies with the standing wave data, and locates the position coordinates of the typical fault in combination with the line information; the complex fault location unit fuses and analyzes the multiple wave heads and the standing wave steady-state characteristics, inputs the trained hybrid model, and outputs the position coordinates of the complex fault; Multi-fault correlation analysis module: including time relationship analysis unit, space relationship analysis unit and correlation comprehensive judgment unit; wherein, the time relationship analysis unit analyzes the arrival time of each wave head, and identifies the time sequence between different fault points; the space relationship analysis unit uses the space-time attention mechanism and sensor position information to evaluate the propagation characteristics of the fault signal in space, and judges whether there is a spatial propagation or influence relationship between the fault points; the correlation comprehensive judgment unit comprehensively analyzes the results in the time and space dimensions, judges the mutual influence between multiple fault points, and obtains the fault correlation conclusion; Fault integration and decision support module: including fault information integration unit and decision assistance unit; wherein, the fault information integration unit integrates the fault type, fault location and correlation information obtained by the fault type identification module, fault location module and multi-fault correlation analysis module; the decision assistance unit provides complete fault information for the operation and maintenance personnel based on the integrated fault information, and proposes preliminary maintenance strategy suggestions.

[0015] Compared with the prior art, the beneficial effects of the present application are: 1、The application adopts a hybrid model of convolutional neural network and long short-term memory network and a space-time attention mechanism, and deeply fuses the space-time features of traveling wave and standing wave data.

[0016] 2、Based on the deep fusion features, the fault type is identified and located, the typical fault is identified in combination with the traveling wave time difference and the standing wave zero sequence characteristics, the location is dependent on the traveling wave time difference and supplemented by the standing wave verification; the complex fault is identified and located by fusing and analyzing the multiple wave heads and the standing wave reflection characteristics.

[0017] 3、For the multi-fault point scene, the hybrid model is used to analyze the time sequence and spatial propagation correlation of the fault signal. BRIEF DESCRIPTION OF DRAWINGS

[0018] Fig. 1 The application is applied to the steps of the artificial intelligence analysis method for traveling wave fault location; Fig. 2 The application is applied to the system structure diagram of the artificial intelligence analysis system for traveling wave fault location. DETAILED DESCRIPTION

[0019] With reference to the accompanying drawings, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all the other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of the present application.

[0020] Embodiment: As shown in the figure, the present application provides a technical solution, Figs. 1-2 As shown in the figure, Fig. 1 The steps of the artificial intelligence analysis method applied to the traveling wave fault location are shown in the schematic diagram. The present application provides an artificial intelligence analysis method applied to the traveling wave fault location, which comprises the following steps: Step S100: Collecting traveling wave data and standing wave data through current sensors and voltage sensors, and performing pretreatment; Specifically, current sensors and voltage sensors are deployed at key nodes of the transmission line to measure the traveling wave transient current and voltage signals at the key nodes, and to collect the voltage and current signals under steady state and quasi-steady state after the fault; The collected original data is pretreated, data alignment and synchronization verification are performed, the time alignment of data collected by different sensor nodes is adjusted by using GPS time stamp to make the data synchronized in time dimension; digital filtering technology is used for data denoising and filtering to remove high-frequency noise and power frequency interference; wavelet transform method is used to analyze the traveling wave signal, extract the instantaneous amplitude, and highlight the key features and instantaneous changes related to the fault in the signal; data loss and abnormal points are processed, and the mean filling method is used for correction.

[0021] ​In a specific embodiment, the traveling wave data and standing wave data are collected by current sensors and voltage sensors, and preprocessed; specifically, the current sensors and voltage sensors are deployed at key nodes of the power transmission line, at positions 5 kilometers, 15 kilometers, and 25 kilometers away from the substation, to measure the traveling wave transient current and voltage signals at the key nodes, and to collect the voltage and current signals under steady state and quasi-steady state after a fault. In a primary line fault, the sensor at the 5-kilometer position collects an initial traveling wave current peak of about 10 kA and a voltage peak of about 200 kV, and then records a series of attenuated oscillation traveling wave signals within 5 milliseconds. At the same time, under steady state, the sensor records a current effective value of about 200 A and a voltage effective value of about 500 kV. The collected raw data is preprocessed, data alignment and synchronization verification are performed, the time alignment of the data collected by different sensor nodes is adjusted using GPS timestamps to synchronize the data in the time dimension, digital filtering technology is used for data denoising and filtering to remove high-frequency noise with a frequency higher than 1 MHz and 50 Hz power frequency interference, db4 wavelet basis function is used for 5-layer decomposition to analyze the traveling wave signals and extract the instantaneous amplitude of the 10-kA peak detected at 5 milliseconds, and the average current of 200 A of the normal nodes in a certain time window is used to fill in the missing data of a node in the certain time window.

[0022] Step S200: Extracting spatio-temporal features from the preprocessed traveling wave and standing wave data using a hybrid model combining a convolutional neural network and a long short-term memory network, and fusing a spatio-temporal attention mechanism to mine the correlation and complementarity between the two modal features and realize deep fusion of information. Specifically, the hybrid model includes a convolutional neural network CNN branch and a long short-term memory network LSTM branch, the preprocessed data is used as the input of the hybrid model, the CNN branch captures the structured information of the signal in the local area, the convolutional layer operates on the time series of the signal through a sliding filter to automatically learn and extract the local spatial features in the signal, and the pooling layer is used to reduce the feature dimension and enhance the translation invariance of the hybrid model; the LSTM branch captures the evolution features of the signal in the time dimension, processes the time sequence dynamics in the sequence data, and learns and remembers the dependency relationships and dynamic patterns in the long time sequence through a gating mechanism, including an input gate, a forgetting gate, and an output gate. After the two branches extract spatial features and time sequence features respectively, the outputs of the two branches are fused and input into a spatio-temporal attention mechanism module, the weights are dynamically calculated and allocated according to the importance of the features for fault type identification and positioning; the importance of the features in the time axis and the importance of the spatial dimension are evaluated, so that the hybrid model can automatically focus on the most relevant spatio-temporal region, suppress noise and redundant information, and effectively mine the correlation and complementarity between the traveling wave signals and the standing wave signals.

[0023] In a specific embodiment, the hybrid model includes a convolutional neural network (CNN) branch and a long short-term memory (LSTM) branch, and the preprocessed data is taken as the input of the hybrid model, each sample including 1000 time steps and 8 features, 4 of which are from the traveling wave and 4 are from the standing wave. The CNN branch captures the structured information of the signal in the local area, and the convolution layer operates on the time series of the signal through a sliding filter with a size of 5 and a step of 1 to extract the peak features reflecting the moment of fault occurrence; the pooling layer adopts maximum pooling with a pooling window size of 2 to reduce the feature dimension and enhance the translation invariance of the hybrid model. The LSTM branch captures the evolution features of the signal in the time dimension and processes the time sequence dynamics in the sequence data through the gating mechanism of its input gate, forget gate and output gate to learn and remember the dependency and dynamic pattern in 100 time steps; the number of LSTM units is set to 64.

[0024] After the spatial features and the time sequence features are extracted in the two branches respectively, the outputs of the two branches are spliced to obtain the fused feature dimension. The spliced features are input into a spatio-temporal attention mechanism module, which includes a shared weight query, key and value generation network, and uses a fully connected layer to map the features to a space with a dimension of 64. By calculating the dot product of the query and the key and passing it through a softmax function, the attention weight of each time step and spatial position is obtained to evaluate the importance of the features in the time axis and the importance of the spatial dimension, so that the hybrid model can automatically focus on the most relevant spatio-temporal region, including the moment of fault occurrence and the subsequent few time steps, as well as the features corresponding to the region near the fault point, to suppress noise and redundant information and effectively mine the correlation and complementarity between the initial traveling wave head features and the reflected wave features after the steady state.

[0025] Step S300: Based on the fused features, the hybrid model is used for fault type identification and fault location; the fault type of the typical fault is obtained by combining the traveling wave grounding feature and the zero sequence component identification in the standing wave; the information dimension provided by the feature combination of the traveling wave and the standing wave is fused and analyzed to obtain the fault type of the complex fault; Specifically, the fused features are used in combination with the hybrid model to identify and locate the fault type; in the fault type identification link, the hybrid model focuses on analyzing the traveling wave grounding transient characteristics and the zero sequence component information in the standing wave reflected in the fused features to distinguish typical fault types; the typical fault types include single-phase grounding, two-phase short-circuit grounding, two-phase short-circuit and three-phase short-circuit; In identifying single-phase ground fault, the initial transient signal of the traveling wave generated by the single-phase ground fault is captured, which is manifested as that the fault-phase voltage is lower than the preset single-phase ground fault phase voltage threshold, the non-fault-phase voltage is higher than the preset single-phase ground fault non-fault-phase voltage threshold, and the initial traveling wave current of the fault phase is greater than the preset single-phase threshold; the hybrid model can identify the initial disturbance characteristics of the single-phase asymmetry, including the difference in amplitude and polarity of the fault phase and the non-fault phase traveling wave; during the steady state or quasi-steady state after the fault, the single-phase ground fault generates zero sequence current and zero sequence voltage higher than the preset single-phase ground fault zero sequence threshold, and the hybrid model analyzes the frequency spectrum or time domain information of the standing wave part of the fusion feature to detect the existence and characteristics of the zero sequence component; since the zero sequence component has a specific amplitude and mode when single-phase grounding occurs; when the initial traveling wave single-phase asymmetry disturbance is detected, and it is confirmed that the zero sequence component in the standing wave exists in the single-phase ground mode, it is determined as single-phase ground fault; In identifying two-phase short-circuit ground fault, the initial transient signal of the traveling wave of the two-phase short-circuit and at least one-phase grounding is captured; the traveling wave signal is manifested as that the two fault-phase voltages are lower than the respective preset two-phase short-circuit ground fault phase voltage threshold, and the non-fault-phase voltage is higher than or equal to the preset two-phase short-circuit ground fault non-fault-phase voltage threshold; the hybrid model analyzes the initial transient signal of the traveling wave, learns and identifies the specific characteristics of the phase relationship of the two fault-phase currents under this fault, which is manifested as that the initial phase difference is within the preset two-phase phase difference threshold interval; the two phases generate initial traveling wave currents with amplitudes exceeding the preset two-phase threshold, and there is a specific phase relationship; two-phase short-circuit grounding generates zero sequence current and zero sequence voltage, but their size and characteristics are different from single-phase grounding, and the zero sequence component here is caused by two-phase grounding or one-phase grounding and another phase short-circuit, which has identifiable amplitude and phase characteristics; when the two-phase asymmetry disturbance of the traveling wave signal is detected, and it is confirmed that the zero sequence component in the standing wave exists in the two-phase grounding mode, it is determined as two-phase short-circuit ground fault; In identifying two-phase short-circuit fault, the hybrid model analyzes the initial transient characteristics of the traveling wave signal, which is manifested as that the two-phase voltages and currents involved are lower than or deviate from the preset two-phase short-circuit fault voltage and current normal range, and the voltage and current of the non-fault phase are maintained within the normal range; since there is no grounding, when the hybrid model analyzes the fusion feature of the standing wave part, it will detect the zero sequence component with an amplitude lower than the preset two-phase short-circuit fault zero sequence threshold or equal to zero, which is also a key feature for distinguishing; when the hybrid model detects the initial transient characteristics of the two-phase disturbance in the traveling wave signal, and confirms that the amplitude of the zero sequence component in the standing wave is lower than the preset two-phase short-circuit fault zero sequence threshold, it is determined as two-phase short-circuit fault; In identifying a three-phase short-circuit fault, the hybrid model analyzes the initial transient characteristics of the three-phase simultaneous short-circuit reflected in the fusion features, the traveling wave signals are below the preset three-phase short-circuit fault voltage threshold of each phase, the three-phase currents also deviate from the preset three-phase short-circuit fault current normal range, and there is a specific phase relationship between the three phases, which is that the initial change start time difference of the three-phase voltage and current is less than the preset three-phase time threshold, and the change waveform similarity is higher than the preset three-phase similarity threshold; the hybrid model analyzes the fusion features of the standing wave part, and detects the zero sequence component with an amplitude lower than the preset three-phase short-circuit fault zero sequence threshold or equal to zero; when the hybrid model detects the initial transient characteristics of the three-phase simultaneous disturbance in the traveling wave signal and confirms that the amplitude of the zero sequence component in the standing wave is lower than the preset three-phase short-circuit fault zero sequence threshold, it is determined as a three-phase short-circuit fault.

[0026] The complex fault types include high-resistance ground fault and intermittent fault; In identifying a high-resistance ground fault, the hybrid model analyzes the fusion features that the amplitude of the initial transient signal of the traveling wave is lower than the preset high-resistance ground fault traveling wave amplitude threshold, the waveform change exceeds the preset high-resistance ground wave waveform normal range, and the transient signal is accompanied by noise or oscillation with an amplitude lower than the preset noise threshold and a frequency higher than the preset high-frequency threshold; at the same time, in the standing wave analysis, the amplitude of the zero sequence component is lower than that of the typical ground fault, the rate of change with time is lower than the preset zero sequence component change threshold, and the fluctuation exceeds the preset high-resistance ground zero sequence component fluctuation normal range; the hybrid model determines it as a high-resistance ground fault by analyzing the characteristic combination that the traveling wave transient is lower than the preset high-resistance ground fault traveling wave amplitude threshold and the amplitude of the standing wave zero sequence component is low and unstable; In identifying an intermittent fault, the hybrid model analyzes the time-varying nature of the traveling wave signal in the fusion features, which is that the appearance interval of the traveling wave transient signal of the fault phase voltage and current exceeds the preset minimum interval time, the disappearance interval exceeds the preset maximum disappearance time, the amplitude exceeds the preset intermittent fault amplitude range, and the polarity change exceeds the preset polarity rule change range, and the arrival time interval stability is lower than the preset intermittent fault interval stability threshold; at the same time, in the standing wave analysis, the existence time of the zero sequence component is lower than the preset intermittent fault zero sequence component duration threshold; the hybrid model determines it as an intermittent fault by capturing the characteristic combination that both the traveling wave transient and the standing wave zero sequence component do not meet the preset continuity or stability and have obvious intermittent characteristics.

[0027] In a specific embodiment, the hybrid model first analyzes the initial transient signal of the traveling wave, the transient signal amplitude 4.5kA is significantly higher than the preset ground fault traveling wave amplitude threshold 2kA, and the waveform contains a main peak with an amplitude exceeding the preset noise threshold 100A and a frequency higher than the preset high-frequency threshold 5kHz. The peak appears as a pulse with a frequency of about 10kHz and an amplitude of 3kA in the time domain waveform. At the same time, in the standing wave analysis, the hybrid model observes that the maximum amplitude reaches 250A, which is much higher than the preset typical ground fault zero sequence component threshold 200A; its rate of change over time is also higher than the preset zero sequence component change threshold 20A / s, and the actual rate of change is as high as 35A / s; and during the fault duration of 0.1 to 0.5 seconds, the zero sequence current fluctuates relatively stably between 200A and 300A, and the fluctuation amplitude does not exceed the preset 50A. The hybrid model determines the fault type to be single-phase ground fault by comprehensively analyzing this unique combination of characteristics, i.e., the traveling wave transient amplitude is high and the waveform is dramatic, while the standing wave zero sequence component amplitude is high, the change rate is high, and the fluctuation is moderate.

[0028] The hybrid model first analyzes the initial transient signal of the traveling wave, the amplitude 800A is lower than the preset high-resistance ground fault traveling wave amplitude threshold 1kA, and the waveform contains multiple noise or oscillation components with an amplitude exceeding the preset noise threshold 50A but a frequency higher than the preset high-frequency threshold 1kHz. These components appear as burrs with a frequency of about 2kHz and an amplitude fluctuating between ±30A in the time domain waveform. At the same time, in the standing wave analysis, the hybrid model observes that the maximum amplitude is only 100A, which is much lower than the preset typical ground fault zero sequence component threshold 200A; its rate of change over time 5A / s is also lower than the preset zero sequence component change threshold 10A / s; and during the fault duration of 0.1 to 0.3 seconds, the zero sequence current fluctuates irregularly between 70A and 130A, and the fluctuation amplitude exceeds the preset 20A. The hybrid model determines the fault type to be high-resistance ground fault by comprehensively analyzing this unique combination of characteristics, i.e., the traveling wave transient amplitude is low and the waveform is abnormal, while the standing wave zero sequence component amplitude is low, the change rate is low, and the fluctuation is large.

[0029] Step S400: In terms of fault location, for typical faults, the time difference between the arrival of the initial traveling wave at the two end measurement points is focused on, and the fault distance is calculated in combination with the line parameters, and the standing wave information is used as auxiliary verification; for complex faults, the time interval and energy distribution of multiple wave heads are analyzed, and the fault location is comprehensively judged by relying on the reflection characteristics of the fault point reflected in the standing wave information; Specifically, after identifying that the fault type is a typical fault, the time difference T d between the arrival of the initial traveling wave at the two end measurement points is used to calculate the fault distance. ; wherein, D is the fault distance, representing the distance value from one end of the line to the fault point, c is the wave speed, T d is the time difference of the initial traveling wave head reaching the two ends of the line measurement points; In this process, the steady-state features extracted from the standing wave data are used as auxiliary verification information, which are compared with the expected standing wave reflection characteristics corresponding to the fault distance D calculated based on the traveling wave, to verify the traveling wave positioning result and improve the confidence of the positioning result. The fault distance D is output, and the mapping relationship between the fault distance D and the fault positioning coordinate point is established combined with the geographical or topological information of the line. Through linear interpolation calculation, the specific coordinate point of the fault positioning is obtained.

[0030] When the fault type is identified as a complex fault, the mixed model analyzes the time interval sequence and corresponding energy distribution features of multiple traveling wave heads reaching the measurement points extracted from the traveling wave data; and analyzes the steady-state features extracted from the standing wave data, including the evolution law of the standing wave ratio and the reflection coefficient with frequency and time; Based on the comprehensive analysis of the multiple wave head time interval, energy distribution and standing wave reflection characteristics, a unified multi-dimensional feature vector is formed, and the mixed model establishes the mapping relationship between the fault location and the multi-dimensional feature vector. Using historical fault data, the mixed model learns the feature vectors generated by high resistance grounding and intermittent faults at different locations. Through training, the mixed model automatically learns and fits the nonlinear relationship between the feature vector and the fault location. The extracted feature vector is input into the trained mixed model, and the specific fault location coordinate value is output through calculation. The coordinate value is the final determined fault positioning coordinate, realizing the location positioning of high resistance grounding and intermittent faults.

[0031] In a specific embodiment, after identifying the fault type as single-phase grounding, the time difference T d is 2.4 ms, the line wave speed c is 285 m / s, and the calculated fault distance D is 342 m. In this process, the steady-state features extracted from the standing wave data are used as auxiliary verification information. The expected standing wave ratio at the fault distance of 342 m is greater than 1.5, and the amplitude of the reflection coefficient is greater than 0.2. These features are used as auxiliary verification information. The actual extracted standing wave features are compared with the expected standing wave reflection characteristics corresponding to the fault distance of 342 m calculated based on the traveling wave. It is found that they are consistent, thereby verifying the accuracy of the traveling wave positioning result and improving the confidence of the positioning result. After outputting the fault distance D=342 m, combined with the known total length of the line of 10 km, the starting coordinate (0, 0), and the ending coordinate (10 km, 0), the mapping relationship between the fault distance D and the fault positioning coordinate point is established. Through linear interpolation calculation, the fault point is located at a distance of 3.42 km from the starting point, and the coordinate point is (3.42 km, 0).

[0032] When the fault type is identified as a high-resistance ground fault, the hybrid model analyzes the time interval sequence and corresponding energy distribution characteristics of multiple wave heads extracted from the traveling wave data reaching the measurement point. The second wave head appears 0.8 ms after the first wave head, with an energy of 30% of the first; the third wave head appears 1.2 ms after the second wave head, with an energy of 20% of the first. The steady-state characteristics extracted from the standing wave data are analyzed, and the reflection coefficient amplitude reaches a peak of 0.35 at 100 Hz, and presents periodic fluctuations over time. Based on the comprehensive analysis of the time interval, energy distribution and standing wave reflection characteristics of multiple wave heads, a unified multi-dimensional feature vector is formed. The hybrid model uses historical fault data to learn the feature vectors generated by high-resistance ground faults and intermittent faults at different locations. Through training, the nonlinear relationship between the feature vector and the fault location is automatically learned and fitted. The extracted feature vector is input into the trained hybrid model, and the coordinates (1.2 km, 3.5 km) are obtained by calculation. The coordinate value is the final determined fault location coordinates.

[0033] Step S500: When multiple fault points are detected near the fault location, a correlation analysis is performed using the hybrid model to evaluate whether there is a time sequence and spatial propagation relationship between the fault signals, and the fault correlation is obtained. Specifically, when multiple fault points exist, the hybrid model analyzes the arrival time of each wave head to identify the time relationship between different fault points and determine the propagation order of the fault signals. The hybrid model uses a spatio-temporal attention mechanism to evaluate the spatial propagation characteristics of the fault signals and determine whether there is a spatial propagation relationship between the fault points, including signal reflection, refraction and scattering phenomena. In terms of time analysis, the hybrid model analyzes the time difference of fault wave heads to infer the occurrence order of fault points and help locate the sequence of multiple fault points. In terms of spatial analysis, the hybrid model combines the positions of sensor nodes to evaluate the propagation path of signals, fault distance and fault location, and further confirms the spatial correlation of different fault points. Based on the correlation analysis in time and space dimensions, the hybrid model comprehensively judges the mutual influence between multiple fault points and obtains the fault correlation.

[0034] In a specific embodiment, in a complex scenario involving three fault points F1, F2, and F3, sensor A detects the first wave front at timestamp T0 = 0.5 ms, sensor B detects the second wave front at T1 = 1.2 ms, and sensor C detects the third wave front at T2 = 2.0 ms. The hybrid model analyzes these arrival times and infers that the possible sequence of fault occurrences is F1 first, which generates a traveling wave that propagates to A, B, and C, followed by F2, whose wave front is detected by B and C, and finally F3, whose wave front is mainly detected by C. The hybrid model uses a spatio-temporal attention mechanism to evaluate the propagation characteristics of the fault signals in space, determining whether there is a spatial propagation relationship between the fault points, including signal reflection, refraction, and scattering phenomena. In terms of temporal analysis, the hybrid model analyzes the time differences of the wave fronts, T1-T0 = 0.7 ms and T2-T1 = 0.8 ms, and combines the line length and wave speed c = 285 m / s to infer the sequence of fault point occurrences and relative distances, helping to locate the sequence of multiple fault points. In terms of spatial analysis, the hybrid model combines the known locations of sensor nodes A (0, 0), B (5 km, 0), and C (10 km, 0) to determine that F1 is closest to point A, F2 is near point B, and F3 is near point C, and that the signal of F2 may have been affected by the reflected wave of F1. It determines whether the occurrence of F2 is related to the persistence of F1 or whether F3 is a secondary fault caused by F1 and F2, and ultimately concludes that "F1 is the main fault, and F2 and F3 are secondary faults affected by it."

[0035] Step S600: Based on fault type, fault location, and fault correlation, assisting operation and maintenance personnel in decision-making and diagnosis.

[0036] Specifically, based on the deep analysis of the hybrid model on traveling wave signals and standing wave signals, the fault type is accurately identified, including single-phase grounding, two-phase short-circuit grounding, two-phase short-circuit, three-phase short-circuit, high-resistance grounding, and intermittent fault; by feeding back the identified fault type to the operation and maintenance personnel, the nature of the fault is provided, and corresponding processing measures are determined; Combined with fault type and signal propagation analysis, the hybrid model provides fault location information to determine the specific location of the fault occurrence; by analyzing the propagation time difference of the traveling wave signal, the zero sequence component in the standing wave, and the wave front arrival time, the operation and maintenance personnel are helped to locate the fault point in the power line; When there are multiple fault points, the time and spatial relationship between the fault points is identified through spatio-temporal correlation analysis; the hybrid model analyzes the mutual influence between the fault points, determines whether the fault is independent or has a linkage effect, and assists the operation and maintenance personnel in decision-making according to the fault correlation; Based on fault type, location, and correlation, decision support and maintenance strategies are provided for operation and maintenance personnel.

[0037] In a specific embodiment, the model analyzes the current signal characteristics and determines that the fault type is a high-impedance grounding fault. For high-impedance grounding, the maintenance personnel will prefer to use a high-impedance grounding detector for review. The model calculates that the fault distance from substation A is 3.42 kilometers, with a positioning accuracy of ±50 meters. By analyzing the propagation time difference T d =2.4ms of the traveling wave signal, the zero sequence component amplitude 0.3p.u. in the standing wave, and the wave head arrival time, the maintenance personnel can locate the fault point in the power line. When there are multiple fault points, the fault correlation helps the maintenance personnel make decisions, and determines that the occurrence of F2 may be affected by the insulation damage extension caused by F1. Based on the fault type being a high-impedance grounding fault, the location (3.42km, 0) and (7.89km, 0), and the correlation F2 being caused by F1, the maintenance personnel are provided with decision support and maintenance strategies, and are advised to prioritize F1 and check the area near F1 for potential hidden dangers that may cause F2, thereby more efficiently and accurately developing a repair plan.

[0038] As Fig. 2 The system structure diagram of the artificial intelligence analysis system applied to traveling wave fault location is shown, and the present application provides an artificial intelligence analysis system applied to traveling wave fault location, which comprises: A data acquisition and preprocessing module: including a data acquisition unit and a data preprocessing unit; wherein the data acquisition unit acquires traveling wave data and standing wave data through current sensors and voltage sensors; the data preprocessing unit aligns and synchronizes the data, removes noise and filters, extracts features, and corrects missing and abnormal values; A deep feature extraction and fusion module: including a hybrid model structure unit, a feature extraction unit, and a feature fusion and attention mechanism unit; wherein the hybrid model structure unit constructs a hybrid network structure containing a convolutional neural network CNN and a long short-term memory network LSTM, which are respectively used to process spatial features and time series features; the feature extraction unit extracts spatial structure features of the input data using the CNN branch and extracts time evolution features using the LSTM branch; the feature fusion and attention mechanism unit fuses spatial and temporal features into spatio-temporal features, and applies an attention mechanism to dynamically emphasize important features and suppress irrelevant information; A fault type identification module: including a typical fault identification unit and a complex fault identification unit; wherein the typical fault identification unit focuses on analyzing traveling wave grounding transient characteristics and standing wave zero sequence component information based on the fused features of the spatio-temporal attention mechanism, and identifies typical fault types including single-phase grounding, two-phase short-circuit grounding, two-phase short-circuit, and three-phase short-circuit; the complex fault identification unit analyzes the feature combinations exhibited by traveling waves and standing waves based on the fused features of the spatio-temporal attention mechanism, and identifies high-impedance grounding faults and intermittent faults; The fault positioning module comprises a typical fault positioning unit and a complex fault positioning unit; wherein the typical fault positioning unit calculates the fault distance based on the traveling wave time difference, verifies the typical fault position coordinates by using the standing wave data in combination with the line information; the complex fault positioning unit fuses and analyzes the multiple wave heads of the traveling wave and the steady-state characteristics of the standing wave, inputs the trained hybrid model, and outputs the position coordinates of the complex fault; The multi-fault correlation analysis module comprises a time relationship analysis unit, a space relationship analysis unit and a correlation comprehensive judgment unit; wherein the time relationship analysis unit analyzes the arrival time of each wave head and identifies the time sequence between different fault points; the space relationship analysis unit uses the space-time attention mechanism and the sensor position information to evaluate the spatial propagation characteristics of the fault signals and judge whether there is a spatial propagation or influence relationship between the fault points; the correlation comprehensive judgment unit comprehensively analyzes the results in the time and space dimensions, judges the mutual influence between the multiple fault points, and obtains the fault correlation conclusion; The fault integration and decision support module comprises a fault information integration unit and a decision assistance unit; wherein the fault information integration unit integrates the fault type, fault position and correlation information obtained by the fault type identification module, fault positioning module and multi-fault correlation analysis module; the decision assistance unit provides complete fault information for the operation and maintenance personnel based on the integrated fault information and proposes a preliminary maintenance strategy suggestion.

[0039] It is apparent to those skilled in the art that the present application is not limited to the details of the foregoing exemplary embodiments, and that the present application can be implemented in other particular forms without departing from the spirit or essential characteristics of the present application. Therefore, the embodiments should be considered in all respects as illustrative and not restrictive, and the scope of the present application should be defined by the appended claims rather than the above description, and it is intended to include all changes falling within the meaning and scope of equivalents of the claims. Any reference signs in the claims should not be considered as limiting the claims involved.

Claims

1. An artificial intelligence analysis method applied to traveling wave fault location, characterized in that, include: Traveling wave data and standing wave data are collected using current sensors and voltage sensors, and then preprocessed. By using a hybrid model combining convolutional neural networks and long short-term memory networks, spatiotemporal features are extracted from preprocessed traveling wave and standing wave data. A spatiotemporal attention mechanism is incorporated to explore the correlation and complementarity between the two modal features, thereby achieving deep information fusion. Based on the fused features, a hybrid model is used for fault type identification and fault location; by combining the characteristics of traveling wave grounding and the identification of zero-sequence components in standing waves, the fault types of typical faults are obtained; by fusing and analyzing the information dimensions provided by the combination of features of traveling wave and standing wave performance, the fault types of complex faults are obtained. In terms of fault location, for typical faults, we focus on the time difference between the arrival of the initial traveling wave front at both measurement points, and calculate the fault distance by combining the line parameters, with standing wave information used as an auxiliary verification. For complex faults, the time intervals and energy distribution of multiple wavefronts are analyzed, and the fault location is determined by comprehensively considering the reflection characteristics of the fault point reflected in the standing wave information. When multiple fault points are detected near the fault location, a hybrid model is used to perform correlation analysis to evaluate whether there is a temporal sequence and spatial propagation relationship between the fault signals, and to obtain the fault correlation. Based on the fault type, fault location, and fault correlation, it assists maintenance personnel in making decisions and diagnosing problems.

2. The artificial intelligence analysis method for traveling wave fault location according to claim 1, characterized in that, The process of acquiring traveling wave and standing wave data through current and voltage sensors, and performing preprocessing, includes: Deploy current and voltage sensors at key nodes of transmission lines to measure traveling wave transient current and voltage signals at key nodes, and collect voltage and current signals under steady-state and quasi-steady-state conditions after a fault. The collected raw data is preprocessed, and data alignment and synchronization verification are performed. GPS timestamps are used to adjust the time alignment of data collected by different sensor nodes to synchronize the data in the time dimension. Digital filtering technology is used to denoise and filter the data to remove high-frequency noise and power frequency interference. Wavelet transform is used to analyze the traveling wave signal, extract the instantaneous amplitude, and highlight the key features and instantaneous changes in the signal that are related to the fault. Data loss and outliers are handled and corrected using the mean filling method.

3. The artificial intelligence analysis method for traveling wave fault location according to claim 1, characterized in that, The method utilizes a hybrid model combining convolutional neural networks and long short-term memory networks to extract spatiotemporal features from preprocessed traveling and standing wave data. It incorporates a spatiotemporal attention mechanism to explore the correlation and complementarity between the two modal features, achieving deep information fusion, including: The hybrid model comprises a Convolutional Neural Network (CNN) branch and a Long Short-Term Memory (LSTM) branch. Preprocessed data is used as input to the hybrid model. The CNN branch captures the structured information of the signal within local regions. Convolutional layers operate on the time series of the signal using sliding filters, automatically learning and extracting local spatial features. Pooling layers reduce feature dimensionality and enhance the translation invariance of the hybrid model. The LSTM branch captures the evolutionary features of the signal over time, processing temporal dynamics in sequence data. It learns and memorizes dependencies and dynamic patterns in long-term sequences through gating mechanisms, including input gates, forget gates, and output gates. After extracting spatial and temporal features from the two branches respectively, the outputs of the two branches are fused and input into the spatiotemporal attention mechanism module. The weights are dynamically calculated and assigned according to the importance of features to fault type identification and localization. The importance of features on the time axis and the spatial dimension are evaluated, so that the hybrid model can automatically focus on the most relevant spatiotemporal region, suppress noise and redundant information, and effectively explore the correlation and complementarity between traveling wave signals and standing wave signals.

4. The artificial intelligence analysis method for traveling wave fault location according to claim 1, characterized in that, The combination of traveling wave grounding characteristics and zero-sequence component identification in standing waves yields typical fault types, including: The features fused through a spatiotemporal attention mechanism are combined with a hybrid model to identify and locate fault types. In the fault type identification stage, the hybrid model focuses on analyzing the transient characteristics of traveling wave grounding and the zero-sequence component information in standing waves reflected in the fused features to distinguish typical fault types. Typical fault types include single-phase grounding, two-phase short-circuit grounding, two-phase short circuit, and three-phase short circuit. When identifying a single-phase ground fault, the initial transient signal of the traveling wave generated by the single-phase ground fault is captured. This is manifested as the voltage of the faulty phase being lower than the preset single-phase ground fault phase voltage threshold, while the voltage of the non-faulty phase is higher than the preset single-phase ground fault non-faulty phase voltage threshold. At the same time, the faulty phase generates an initial traveling wave current with an amplitude exceeding the preset single-phase threshold. The hybrid model can identify this single-phase asymmetric initial disturbance characteristic, including the difference in amplitude and polarity of the traveling waves between the faulty and non-faulty phases. During the steady-state or quasi-steady-state period after the fault, the single-phase ground fault generates a zero-sequence current and zero-sequence voltage higher than the preset single-phase ground zero-sequence threshold. The hybrid model analyzes and fuses the spectral or time-domain information of the standing wave component in the characteristics to detect the existence and characteristics of the zero-sequence component. Since the zero-sequence component has a specific amplitude and mode during a single-phase ground fault, when the single-phase asymmetric disturbance of the initial traveling wave is detected, and the presence of a zero-sequence component conforming to the single-phase ground fault mode in the standing wave is confirmed, it can be determined as a single-phase ground fault. When identifying a two-phase short-circuit-to-ground fault, the system captures the initial transient signal of a traveling wave where both phases are simultaneously short-circuited and at least one phase is grounded. The traveling wave signal is characterized by the voltages of the two faulted phases being lower than their respective preset two-phase short-circuit-to-ground fault phase voltage thresholds, while the voltages of the non-faulted phases are higher than or equal to the preset two-phase short-circuit-to-ground non-faulted phase voltage thresholds. The hybrid model analyzes the initial transient signal of the traveling wave, learns and identifies the specific characteristics of the phase relationship between the currents of the two faulted phases under this fault, which is characterized by the initial phase difference being within the preset two-phase phase difference threshold range. The two phases generate initial traveling wave currents with amplitudes exceeding the preset two-phase thresholds, and there is a specific phase relationship. The two-phase short-circuit-to-ground fault generates zero-sequence current and zero-sequence voltage, but their magnitude and characteristics are different from those of a single-phase ground fault. Here, the zero-sequence component is caused by the simultaneous grounding of both phases or by one phase being grounded and short-circuited with the other phase, and has identifiable amplitude and phase characteristics. When a two-phase asymmetrical disturbance of the traveling wave signal is detected, and it is confirmed that there is a zero-sequence component in the standing wave that conforms to the two-phase ground fault mode, it is determined to be a two-phase short-circuit-to-ground fault. When identifying a two-phase short-circuit fault, the hybrid model analyzes the initial transient characteristics of the traveling wave signal, which are manifested as the voltage and current of the two involved phases being lower than or deviating from the preset normal range of two-phase short-circuit fault voltage and current, while the voltage and current of the non-faulty phases remain within the normal range. Since there is no grounding, when the hybrid model analyzes the fusion characteristics of the standing wave component, it will detect a zero-sequence component with an amplitude lower than or equal to the preset two-phase short-circuit fault zero-sequence threshold, which is also a key feature for differentiation. When the hybrid model detects the initial transient characteristics of two phases being disturbed simultaneously in the traveling wave signal and confirms that the amplitude of the standing wave zero-sequence component is lower than the preset two-phase short-circuit fault zero-sequence threshold, it can be determined as a two-phase short-circuit fault. When identifying three-phase short-circuit faults, the hybrid model analyzes the initial transient characteristics of simultaneous short circuits in the three phases reflected in the fusion features. The traveling wave signal is characterized by three-phase voltages simultaneously being lower than their respective preset three-phase short-circuit fault voltage thresholds, and three-phase currents also deviating from their respective preset normal ranges for three-phase short-circuit fault currents. Furthermore, there is a specific phase relationship between the three phases, manifested as the initial time difference of the initial changes in the three-phase voltage and current being less than the preset three-phase time threshold, and the similarity of the changing waveforms being higher than the preset three-phase similarity threshold. The hybrid model analyzes the fusion features of the standing wave component and detects zero-sequence components with amplitudes lower than or equal to the preset three-phase short-circuit fault zero-sequence threshold. When the hybrid model detects the initial transient characteristics of simultaneous disturbance in the three phases in the traveling wave signal and confirms that the amplitude of the zero-sequence component in the standing wave is lower than the preset three-phase short-circuit fault zero-sequence threshold, it can be determined as a three-phase short-circuit fault.

5. The artificial intelligence analysis method for traveling wave fault location according to claim 1, characterized in that, The information dimensions provided by the fusion analysis of the characteristic combination of traveling wave and standing wave behavior yield the fault types of complex faults, including: Complex fault types include high-resistance grounding faults and intermittent faults; When identifying high-resistance grounding faults, the hybrid model analyzes the following characteristics: the amplitude of the initial transient signal of the traveling wave is lower than the preset high-resistance grounding fault traveling wave amplitude threshold, the waveform change exceeds the preset normal range of the high-resistance grounding traveling wave waveform, and the transient signal is accompanied by noise or oscillation with an amplitude lower than the preset noise threshold and a frequency higher than the preset high-frequency threshold. At the same time, in the standing wave analysis, the amplitude of the zero-sequence component is lower than that of a typical grounding fault, the rate of change with time is lower than the preset zero-sequence component change threshold, and the fluctuation exceeds the preset normal range of high-resistance grounding zero-sequence component fluctuation. The hybrid model determines a high-resistance grounding fault by analyzing this combination of characteristics: the traveling wave transient is lower than the preset high-resistance grounding fault traveling wave amplitude threshold, and the standing wave zero-sequence component amplitude is low and unstable. When identifying intermittent faults, the hybrid model analyzes the time-varying nature of the traveling wave signal in the fusion features. This is manifested in the following: the interval between the appearance of the traveling wave transient signals of the fault phase voltage and current exceeds the preset minimum interval time, the interval between their disappearance exceeds the preset maximum disappearance time, the amplitude exceeds the preset intermittent fault amplitude range, the polarity change exceeds the preset polarity regular change range, and the arrival time interval stability is lower than the preset intermittent fault interval stability threshold. At the same time, in the standing wave analysis, the existence time of the zero-sequence component is lower than the preset intermittent fault zero-sequence component persistence time threshold. The hybrid model determines the fault as intermittent by capturing this combination of features where both the traveling wave transient and the standing wave zero-sequence component exhibit obvious intermittent characteristics and do not meet the preset continuity or stability.

6. The artificial intelligence analysis method for traveling wave fault location according to claim 1, characterized in that, Regarding fault location, for typical faults, the focus is on the time difference between the initial traveling wavefront arriving at both measurement points. The fault distance is calculated by combining this with line parameters, and standing wave information is used as auxiliary verification, including: After identifying the fault type as a typical fault, the time difference T between the initial traveling wave front and the measurement points at both ends of the line, extracted from the traveling wave data, is used. d The formula for calculating fault distance is: ； Where D is the fault distance, representing the distance from one end of the line to the fault point, c is the wave velocity, and T is the wave velocity. d The time difference between the arrival of the initial traveling wavefront at the measurement points at both ends of the line; In this process, the steady-state features extracted from the standing wave data are used as auxiliary verification information. By comparing them with the expected standing wave reflection characteristics corresponding to the fault distance D calculated based on the traveling wave, the traveling wave location results are verified, the confidence of the location results is improved, the fault distance D is output, and the mapping relationship between the fault distance D and the fault location coordinates is established by combining the geographical or topological information of the line. The specific coordinates of the fault location are obtained by linear interpolation calculation.

7. The artificial intelligence analysis method for traveling wave fault location according to claim 1, characterized in that, For complex faults, the analysis of time intervals and energy distribution of multiple wavefronts, relying on the reflection characteristics of the fault point reflected in the standing wave information, comprehensively determines the fault location, including: When the fault type is identified as a complex fault, the hybrid model analyzes the time interval sequence of multiple traveling wave fronts arriving at the measurement point and the corresponding energy distribution characteristics extracted from the traveling wave data; it also analyzes the steady-state characteristics extracted from the standing wave data, including the evolution of the standing wave ratio and reflection coefficient with frequency and time. Based on a comprehensive analysis of multiple wavefront time intervals, energy distributions, and standing wave reflection characteristics, a unified multidimensional feature vector is formed. The hybrid model establishes a mapping relationship between the fault location and the multidimensional feature vector. Using historical fault data, it learns the feature vectors generated by high-resistance grounding and intermittent faults at different locations. Through training, the hybrid model automatically learns and fits the nonlinear relationship between the feature vector and the fault location. The currently extracted feature vector is input into the trained hybrid model, and through calculation, the specific fault location coordinates are output. These coordinates are the final determined fault location coordinates, realizing the location of high-resistance grounding and intermittent faults.

8. The artificial intelligence analysis method for traveling wave fault location according to claim 1, characterized in that, When multiple fault points are detected near the fault location, a hybrid model is used for correlation analysis to evaluate whether there is a temporal sequence and spatial propagation relationship between the fault signals, thus obtaining fault correlation, including: When multiple fault points exist, the hybrid model analyzes the arrival time of each wavefront to identify the temporal relationship between different fault points and determine the propagation sequence of the fault signal. The hybrid model utilizes a spatiotemporal attention mechanism to evaluate the spatial propagation characteristics of fault signals and determine whether there are spatial propagation relationships between various fault points, including signal reflection, refraction, and scattering phenomena. In terms of time analysis, the hybrid model analyzes the time difference of the fault wavefront, infers the order of occurrence of fault points, and helps to locate the sequential relationship between multiple fault points; in terms of spatial analysis, the hybrid model combines the location of sensor nodes to evaluate the signal propagation path, fault distance and fault location, and further confirms the spatial correlation of different fault points. Based on the correlation analysis of time and space dimensions, the hybrid model comprehensively judges the mutual influence between multiple fault points and obtains the fault correlation.

9. The artificial intelligence analysis method for traveling wave fault location according to claim 1, characterized in that, The method, based on fault type, fault location, and fault correlation, assists maintenance personnel in decision-making and diagnosis, including: Based on the in-depth analysis of traveling wave and standing wave signals using a hybrid model, fault types are accurately identified, including single-phase grounding, two-phase short-circuit grounding, two-phase short circuit, three-phase short circuit, high-resistance grounding, and intermittent faults. By feeding back the identified fault types to maintenance personnel, the nature of the fault is provided, and corresponding handling measures are determined. By combining fault type and signal propagation analysis, the hybrid model provides fault location information to determine the specific location of the fault; by analyzing the propagation time difference of the traveling wave signal, the zero-sequence component in the standing wave, and the wavefront arrival time, it helps maintenance personnel locate the fault point in the power line. When multiple fault points exist, spatiotemporal correlation analysis is used to identify the temporal and spatial relationships between the fault points; hybrid model analysis is used to analyze the mutual influence between the fault points to determine whether the fault occurs alone or has a linkage effect, and the fault correlation is used to assist maintenance personnel in making decisions. Based on fault type, location, and correlation, it provides decision support and maintenance strategies for operations and maintenance personnel.

10. An artificial intelligence analysis system for traveling wave fault location, using the artificial intelligence analysis method for traveling wave fault location as described in any one of claims 1-9, characterized in that, include: Data acquisition and preprocessing module: includes a data acquisition unit and a data preprocessing unit; the data acquisition unit acquires traveling wave data and standing wave data through current sensors and voltage sensors; the data preprocessing unit performs alignment and synchronization, noise reduction filtering, feature extraction, and correction of missing and outlier values ​​on the data; The deep feature extraction and fusion module includes a hybrid model structure unit, a feature extraction unit, and a feature fusion and attention mechanism unit. The hybrid model structure unit constructs a hybrid network structure comprising a Convolutional Neural Network (CNN) and a Long Short-Term Memory (LSTM) network, used to process spatial features and temporal series features, respectively. The feature extraction unit uses CNN branches to extract spatial structural features from the input data and LSTM branches to extract temporal evolution features. The feature fusion and attention mechanism unit fuses spatial and temporal features into spatiotemporal features and applies an attention mechanism to dynamically emphasize important features and suppress irrelevant information. Fault type identification module: includes a typical fault identification unit and a complex fault identification unit; the typical fault identification unit, based on the features fused by the spatiotemporal attention mechanism, focuses on analyzing the transient characteristics of traveling wave grounding and the zero-sequence component information of standing wave, and identifies typical fault types including single-phase grounding, two-phase short-circuit grounding, two-phase short circuit, and three-phase short circuit; the complex fault identification unit, based on the features fused by the spatiotemporal attention mechanism, integrates and analyzes the characteristic combination of traveling wave and standing wave behavior to identify high-resistance grounding faults and intermittent faults. The fault location module includes a typical fault location unit and a complex fault location unit. The typical fault location unit calculates the fault distance based on the traveling wave time difference, verifies it with standing wave data, and locates the location coordinates of typical faults by combining line information. The complex fault location unit integrates and analyzes the traveling wave multi-wavehead and standing wave steady-state characteristics, inputs a trained hybrid model, and outputs the location coordinates of complex faults. The multi-fault correlation analysis module includes a time relationship analysis unit, a spatial relationship analysis unit, and a correlation comprehensive judgment unit. The time relationship analysis unit analyzes the arrival time of each wavefront to identify the temporal sequence between different fault points. The spatial relationship analysis unit utilizes spatiotemporal attention mechanisms and sensor location information to evaluate the spatial propagation characteristics of fault signals and determine whether there are spatial propagation or influence relationships between various fault points. The correlation comprehensive judgment unit integrates the analysis results from both time and spatial dimensions to determine the mutual influence between multiple fault points and draw conclusions about fault correlation. The fault integration and decision support module includes a fault information integration unit and a decision support unit. The fault information integration unit integrates fault type, fault location, and correlation information obtained from the fault type identification module, fault location module, and multi-fault correlation analysis module. Based on the integrated fault information, the decision support unit provides maintenance personnel with complete fault information and proposes preliminary maintenance strategy suggestions.

Images