A fault monitoring and management method and system for 35kV wind farm overhead collector lines

Abstract

The present invention provides a fault monitoring and management method and system for a 35kV overhead collector line in a wind farm. The method comprises the following steps: obtaining and simplifying a collector line diagram into a topological structure; obtaining fault traveling wave data through a fault data detection terminal; decoupling the fault traveling wave data using the Karen Boolean transform method; obtaining traveling wave modal component signals in different frequency domains using a modal decomposition algorithm; calculating the signal energy spectrum based on the NTEO energy operator, and calibrating the initial wave head of the fault traveling wave; calculating the fault coefficient based on the topological structure and the initial wave head, and locating the fault point; extracting energy spectrum features, and identifying the fault type through a deep learning model; and generating a monitoring report containing the fault point and type. The present invention has the effect of accurately locating the fault point in the wind farm collector line and accurately predicting the fault type.

Description

Technical Field

[0001] The present invention belongs to the technical field of wind power line fault management, and in particular relates to a fault monitoring and management method and system for a 35kV wind farm overhead collector line. Background Art

[0002] As global demand for clean energy continues to grow, wind power, as an important form of renewable energy, has experienced rapid development in recent years. While the construction of large-scale wind farms has made significant contributions to the transformation of the energy structure, it has also brought a series of technical challenges, particularly in terms of operation and maintenance.

[0003] Collector lines, the "blood vessels" of a wind farm, are responsible for collecting and transmitting the electricity generated by individual wind turbines to substations. Their reliability directly impacts the overall wind farm's power generation efficiency and economic benefits. Traditional methods for monitoring wind farm collector line faults rely primarily on regular manual inspections and simple remote monitoring. While these methods may be adequate for small-scale wind farms, their limitations become increasingly apparent as wind farms expand in size and complexity. However, large-scale wind farms typically cover vast areas, with collector lines stretching tens or even hundreds of kilometers, often located in complex terrain and volatile climates. Manual inspections are not only time-consuming and labor-intensive, but may also be impossible to conduct in severe weather conditions. Therefore, existing techniques often use fault diagnosis methods based on steady-state analysis to replace manual inspections. However, the complex topology of wind farms, with their multiple branch nodes and complex dynamic operating characteristics affected by weather, significantly reduce the accuracy of steady-state analysis fault diagnosis methods, ultimately making it difficult to accurately locate line faults within a wind farm and accurately analyze the fault type. Summary of the Invention

[0004] The present invention provides a fault monitoring and management method and system for a 35kV wind farm overhead collector line, so as to solve the problem that it is difficult to accurately locate the line fault point in the wind farm and accurately analyze the fault type of the line fault.

[0005] In a first aspect, the present invention provides a method for fault monitoring and management of a 35kV wind farm overhead collector line, the method comprising the following steps:

[0006] Obtaining a collection line diagram of a wind farm collection line;

[0007] Simplifying the collector line diagram into a collector line topology structure, wherein the collector line topology structure includes a collector main line directly connected to the wind farm power transmission system, and a plurality of collector branch lines connected to the collector main line;

[0008] Acquiring fault traveling wave data in the wind farm collector line through a fault data detection terminal deployed in the collector main line;

[0009] Decoupling the fault traveling wave data into fault traveling wave components using a Karen Boolean transform method;

[0010] Decomposing the fault traveling wave component by using a modal decomposition algorithm to obtain a plurality of traveling wave modal component signals in different frequency domains;

[0011] The signal energy spectrum of all the traveling wave modal component signals is calculated based on the NTEO energy operator, and the initial wave head of the fault traveling wave of the fault traveling wave data is calibrated according to the signal energy spectrum of the highest frequency traveling wave modal component signal;

[0012] Based on the initial wave head of the fault traveling wave and according to the collector line topology structure, the fault coefficients of all the collector branch lines are calculated, and the line fault point in the collector line topology structure is located by combining all the fault coefficients;

[0013] Extracting energy spectrum features of all the signal energy spectra, and identifying the fault type of the line fault point based on the energy spectrum features and using a deep learning model;

[0014] A fault monitoring report is generated in combination with the line fault point and the fault type.

[0015] Optionally, simplifying the current collection circuit diagram into a current collection circuit topology structure comprises the following steps:

[0016] Identify key components in the current collection circuit diagram and current collection circuits between the key components;

[0017] Simplifying all the key components into key nodes, and simplifying the collection lines into key node edges between the key nodes to form multiple topological lines, wherein the key nodes include wind turbine nodes, collection station nodes, fault data detection terminal nodes, and wind farm transmission system nodes, and the number of the fault data detection terminal nodes is 2;

[0018] adding node edge attributes to the key node edges according to the actual length of the collector line, and adding key node attributes to the wind turbine generator set nodes according to the unit status information of the wind turbine generator set corresponding to the wind turbine generator set nodes;

[0019] For any of the topological lines, if both ends of the topological line are the fault data detection terminal nodes, and all the nodes in the topological line except the fault data detection terminal node are the collector station nodes, then the topological line is marked as a collector main line, and all the collector station nodes are connected to the wind farm transmission system nodes through the collector main line;

[0020] If the two ends of the topological line are the collector station node and the wind turbine generator set node respectively, the topological line is marked as a collector branch line;

[0021] When all the topological lines are marked, the simplified collector line topological structure of the collector line diagram is obtained.

[0022] Optionally, decomposing the fault traveling wave component by using a modal decomposition algorithm to obtain a plurality of traveling wave modal component signals in different frequency domains comprises the following steps:

[0023] Collecting wind farm meteorological information of the same time period as the fault traveling wave data by a weather monitoring module configured in the wind farm transmission system;

[0024] Calculating the weather severity during the period of acquiring the fault traveling wave data based on the wind farm weather information and according to preset weather condition rules;

[0025] Integrating the key node attributes of all the wind turbine generator set nodes, and calculating the unit state complexity of all the wind turbine generator set nodes according to all the key node attributes;

[0026] Calculating the signal stability of the fault traveling wave data by combining the weather severity and the unit state complexity;

[0027] If the signal stability is less than a preset stability threshold, an empirical mode decomposition algorithm is used to decompose the fault traveling wave component to obtain a plurality of traveling wave modal component signals in different frequency domains;

[0028] If the signal stability is greater than or equal to the stability threshold, a variational mode decomposition algorithm is used to decompose the fault traveling wave component to obtain a plurality of traveling wave modal component signals in different frequency domains.

[0029] Optionally, the step of decomposing the fault traveling wave component by using an empirical mode decomposition algorithm to obtain a plurality of traveling wave modal component signals in different frequency domains comprises the following steps:

[0030] Calculating the signal maximum and signal minimum of the fault traveling wave component, combining the signal maximum and signal minimum and using a cubic spline interpolation method to obtain the upper envelope and lower envelope of the fault traveling wave component;

[0031] Performing an average calculation on the upper envelope and the lower envelope to obtain an average component of the fault traveling wave component;

[0032] taking the difference between the fault traveling wave component and the average component as a preliminary IMF component;

[0033] If the number of zero points and the number of extreme values ​​of the prepared IMF component are equal, or the average values ​​of the upper and lower envelopes of the prepared IMF component at any time are both 0, then the prepared IMF component is used as a traveling wave modal component signal in one of the frequency domains;

[0034] Calculating the remaining signal component after subtracting the traveling wave modal component signal from the fault traveling wave component;

[0035] The remaining signal components are used as a new round of fault traveling wave components and the above component decomposition steps are repeated until a component decomposition cutoff condition is met;

[0036] After completing the component decomposition of the fault traveling wave component, a soft limit function is used to filter out the signal noise of all the traveling wave modal component signals in the corresponding frequency domain;

[0037] The component decomposition cutoff condition is expressed as follows:

[0038]

[0039] Where: M represents the number of decomposed traveling wave modal component signals, represents the traveling wave modal component signal obtained in the mth round of component decomposition step, represents the traveling wave modal component signal obtained in the m-1th round of component decomposition step, and ε represents the cutoff threshold.

[0040] Optionally, decomposing the fault traveling wave component by using a variational mode decomposition algorithm to obtain a plurality of traveling wave modal component signals in different frequency domains comprises the following steps:

[0041] A decomposition constraint model of the fault traveling wave component is constructed based on the decomposition constraint conditions of the variational modal decomposition algorithm, wherein the decomposition constraint conditions are as follows: N traveling wave modal component signals obtained by decomposing the fault traveling wave component using the variational modal decomposition algorithm have the smallest deviation from the fault traveling wave component after reconstruction, and the sum of the estimated bandwidths of all the traveling wave modal component signals is the smallest;

[0042] Introducing a Lagrangian operator and a penalty factor into the decomposition constraint model to construct a model augmented expression of the decomposition constraint model;

[0043] Based on the model augmented expression and using the alternating direction method of the Lagrangian operator, alternating iterative optimization is performed in the frequency domain until the Wiener filtering of all the traveling wave modal components meets the convergence condition, and each round of alternating iterative optimization process decomposes the traveling wave modal component signal of a different frequency domain from the fault traveling wave component;

[0044] The convergence condition is expressed as follows:

[0045]

[0046] Where: K represents the number of traveling wave modal component signals that have been decomposed, represents the Wiener filtering of the traveling wave modal component signal obtained during the s-th round of alternating iterative optimization, represents the Wiener filtering of the traveling wave modal component signal obtained in the s+1th round of alternating iterative optimization, ||·||2 represents the Euclidean norm, and ζ represents the convergence accuracy.

[0047] Optionally, the steps of calculating the signal energy spectra of all the traveling wave modal component signals based on the NTEO energy operator and calibrating the fault traveling wave initial wave head of the fault traveling wave data according to the signal energy spectrum of the highest frequency traveling wave modal component signal include the following steps:

[0048] converting all of the traveling wave modal component signals into discrete component signals;

[0049] The signal energy spectrum of each discrete component signal is calculated using the NTEO energy operator. The calculation formula of the signal energy spectrum is as follows:

[0050]

[0051] Where: Ψ[h i (x)] represents the i-th discrete component signal h i (x) signal energy spectrum, T represents the resolution parameter, M represents the signal number of the discrete component signal;

[0052] The signal energy spectrum corresponding to the traveling wave modal component signal in the highest frequency domain is selected as the target signal energy spectrum, and the time point with the highest energy value in the target signal energy spectrum is calibrated as the initial wave head of the fault traveling wave data.

[0053] Optionally, the fault traveling wave component includes a fault traveling wave line mode component and a fault traveling wave zero mode component, the fault traveling wave line mode component and the fault traveling wave zero mode component share the fault traveling wave initial wave head, and the fault coefficients of all the collector branch lines are calculated based on the fault traveling wave initial wave head and according to the collector line topology structure, and locating the line fault point in the collector line topology structure in combination with all the fault coefficients includes the following steps:

[0054] The fault data detection terminal node closest to the wind farm power transmission system node in the collector line topology is used as the main line starting node, and the fault data detection terminal node farthest from the wind farm power transmission system node is used as the main line terminating node;

[0055] The direction from the starting node of the main line to the ending node of the main line is defined as the positive direction of the traveling wave;

[0056] Based on the initial wave head of the fault traveling wave and the first propagation velocity of the fault traveling wave line mode component, respectively calculating a first propagation time for the fault traveling wave line mode component to propagate to the starting node of the main line and a second propagation time for the fault traveling wave line mode component to propagate to the ending node of the main line;

[0057] Based on the initial wave head of the fault traveling wave and the second propagation velocity of the zero-mode component of the fault traveling wave, respectively calculating a third propagation time for the zero-mode component of the fault traveling wave to propagate to the starting node of the main line and a fourth propagation time for the zero-mode component of the fault traveling wave to propagate to the ending node of the main line;

[0058] Calculate, by combining the first propagation time, the second propagation time, the third propagation time, the fourth propagation time, and the first propagation speed, a first fault distance between an imaginary fault point and a starting node of the main line, and a second fault distance between the imaginary fault point and a terminating node of the main line;

[0059] Combining the first fault distance, the second fault distance, and the node edge attributes in the collector line topology structure, and calculating according to the positive direction of the traveling wave to obtain the fault coefficients of all the collector branch lines;

[0060] If there is any collector branch line with a fault coefficient of 0, it is determined that the line fault point generating the fault traveling wave data is located on the fault collector branch line with a fault coefficient of 0, and the midpoint of the fault collector branch line is used as the line fault point in the collector line topology structure;

[0061] If all the fault coefficients are not 0, it is determined that the line fault point is located in the main collector line. The initial wave head of the fault traveling wave, the propagation speed of the fault traveling wave component and the node edge attribute are combined for calculation, and the double-terminal traveling wave positioning method is used to determine the line fault point in the collector line topology structure.

[0062] Optionally, the calculation formula of the first fault distance is as follows:

[0063]

[0064] Where: represents the first fault distance between the hypothetical fault point G and the main line starting node Q1, t1 represents the first propagation time, t2 represents the second propagation time, t3 represents the third propagation time, t4 represents the fourth propagation time, and v1 represents the first propagation speed;

[0065] The calculation formula of the second fault distance is as follows:

[0066]

[0067] Where: Indicates the hypothetical fault point G and the main line termination node Q n The second fault distance between

[0068] Optionally, extracting energy spectrum features of all the signal energy spectra and identifying the fault type of the line fault point through a deep learning model based on the energy spectrum features includes the following steps:

[0069] For each of the traveling wave modal component signal's signal energy spectrum, energy spectrum segments of different time lengths are cut out from the signal energy spectrum through time windows of different time scales, and energy time domain features are extracted from each of the energy spectrum segments, wherein the energy time domain features include energy peak value, energy decay rate, energy rise time, energy duration, and energy phase difference feature;

[0070] Extracting energy frequency domain features of the signal energy spectrum in the frequency domain using a statistical method, wherein the energy frequency domain features include energy proportion, energy mean, energy variance and energy kurtosis;

[0071] Using a correlation analysis method, a subset of energy features capable of distinguishing fault types is selected from the energy time domain features and the energy frequency domain features;

[0072] The energy feature subset is input into a pre-trained energy feature recognition model, and the fault type of the line fault point is output through the energy feature recognition model. The energy feature recognition model is a convolutional neural network model including a multi-head attention mechanism.

[0073] In a second aspect, the present invention also provides a fault monitoring and management system for a 35kV wind farm overhead collector line, comprising a memory, a processor, and a computer program stored on the memory and runnable on the processor. When the processor executes the computer program, it implements the fault monitoring and management method for the 35kV wind farm overhead collector line as described in the first aspect.

[0074] The beneficial effects of the present invention are:

[0075] By acquiring and simplifying the collector line diagram into a topological structure, this method lays a solid foundation for subsequent fault location and analysis. This method can better reflect the actual structure of the wind farm collector line and improve fault location accuracy. Advanced signal processing techniques, such as the Karen Boolean transform and modal decomposition algorithm, effectively decouple and decompose fault traveling wave data, thereby extracting more accurate fault signature information. In particular, the method of using the NTEO energy operator to calculate the signal energy spectrum and calibrate the initial wave head of the fault traveling wave significantly improves the accuracy and reliability of fault location. Furthermore, this method innovatively combines the collector line topology with a fault coefficient calculation method. This method not only accurately locates the fault point but also adapts to complex wind farm collector line structures, overcoming the difficulty of traditional methods in locating faults in complex networks. Furthermore, this method also introduces a deep learning model to identify fault types. This method fully leverages big data and artificial intelligence technologies to continuously learn and optimize the fault identification model, improving the accuracy and adaptability of fault type identification. Finally, an accurate fault monitoring report is generated by combining fault location and fault type, providing wind farm operation and maintenance personnel with comprehensive and accurate fault information, which helps to quickly formulate maintenance strategies, reduce downtime, and improve the overall operation efficiency of the wind farm. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 This is a flow chart of a fault monitoring and management method for a 35kV wind farm overhead collector line in one embodiment of the present application.

[0077] Figure 2 This is a schematic diagram of the topological structure of the collector circuit in one embodiment of the present application. DETAILED DESCRIPTION

[0078] The following will be combined with the accompanying drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the embodiments described are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field are within the scope of protection of this application.

[0079] The terms "first," "second," and the like in the specification and claims of this application are used to distinguish similar objects, and are not used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of this application can be implemented in an order other than that illustrated or described herein, and that the objects distinguished by "first," "second," and the like are generally of the same type, and do not limit the number of objects; for example, the first object can be one or more. In addition, the term "and / or" in the specification and claims refers to at least one of the connected objects, and the character " / " generally indicates that the objects connected are in an "or" relationship.

[0080] Figure 1 FIG. 1 is a flow chart of a method for fault monitoring and management of a 35kV wind farm overhead collector line in one embodiment. It should be understood that although Figure 1 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. In addition, Figure 1 At least part of the steps in the above process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but may be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but may be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps. Figure 1 As shown, the present invention discloses a fault monitoring and management method for a 35kV wind farm overhead collector line, which specifically includes the following steps:

[0081] S101. Obtain a collection line diagram of a wind farm collection line.

[0082] This step is accomplished by collecting engineering design drawings, site survey data, and geographic information system (GIS) data for the wind farm. The collector line diagram contains the location information of key equipment such as wind turbines, collector stations, and transmission systems, as well as the direction and length of the overhead lines connecting these devices. Obtaining a complete and accurate collector line diagram is crucial for subsequent fault monitoring and location. In actual operation, professional power system drawing software can be used to draw and store the collector line diagram. This electronic collector line diagram is not only easy to update and manage, but can also be integrated with other systems such as SCADA (Supervisory Control and Data Acquisition) systems to improve overall operation and maintenance efficiency.

[0083] S102. Simplify the collector circuit diagram into a collector circuit topology structure.

[0084] First, the key components of the collection line graph are identified, including wind turbines, collection stations, fault data detection terminals, and the wind farm transmission system. These key components are then simplified into nodes, and the lines connecting these nodes are simplified into edges, forming a network structure consisting of nodes and edges. In this structure, lines directly connected to the wind farm transmission system are labeled as main collection lines, while other lines connected to the main collection lines are labeled as branch collection lines. Each edge is assigned an attribute representing the actual line length, and each node representing a wind turbine is assigned an attribute representing the turbine status. This simplified topology retains the key information of the original collection line while significantly reducing the complexity of subsequent computations. For example, the collection line graph of a wind farm with 50 wind turbines might be simplified into a topology consisting of 53 nodes (50 wind turbine nodes, 2 collection station nodes, and 1 transmission system node) and 52 edges.

[0085] S103. Obtain fault traveling wave data in the wind farm collector line through a fault data detection terminal deployed in the collector main line.

[0086] Among them, the fault data detection terminal is a high-precision measurement device installed on the main collector line. It typically uses GPS clock synchronization technology to accurately capture the instantaneous voltage and current values ​​at the time of the fault. When a line fault occurs, high-frequency electromagnetic transients, known as fault traveling waves, are generated at the fault point. These traveling waves propagate along the line at speeds close to the speed of light and are recorded by the fault data detection terminal. A typical fault data detection terminal has a sampling frequency exceeding 1 MHz, capable of capturing signal changes down to the microsecond level. The acquired fault traveling wave data typically includes the instantaneous values ​​of the three-phase voltage and current, along with the corresponding timestamps. For example, in the case of a single-phase ground fault, the fault data detection terminal may record a sudden drop in phase A voltage and a sharp increase in phase A current, while these changes are not apparent in the other two phases. This high-precision, high-time-resolution fault traveling wave data provides critical information for subsequent fault location and type identification, and serves as the fundamental data source for the entire fault monitoring and management approach.

[0087] S104. Decouple the fault traveling wave data into fault traveling wave components using the Karen Boolean transform method.

[0088] The Karen Boolean transform is a mathematical method that converts three-phase quantities into modal quantities. It can decouple coupled three-phase fault traveling wave data into independent modal components. Specifically, the Karen Boolean transform converts three-phase voltage or current data into three modal components: zero mode, positive mode, and negative mode. The transformation matrix T is usually defined as:

[0089]

[0090] Where a = e(j2π / 3) Assuming the three-phase voltages are Va, Vb, and Vc, the modal voltages are:

[0091]

[0092] Among them, V0 is the zero mode component, V1 is the positive mode component, and V2 is the negative mode component.

[0093] The advantage of this transformation is that it decomposes the complex three-phase coupled fault information into independent modal components, simplifying the subsequent analysis process. For example, in a single-phase ground fault, the zero-mode component will increase significantly, while in a three-phase fault, the positive mode component will dominate. By observing the characteristics of different modal components, the type and severity of the fault can be preliminarily determined. In addition, modal decomposition can reduce the influence of phase mutual inductance, improving the accuracy of fault location. This step lays the foundation for subsequent modal decomposition and feature extraction, and is a key step in achieving accurate fault location and type identification.

[0094] S105. Decompose the fault traveling wave component using a modal decomposition algorithm to obtain multiple traveling wave modal component signals in different frequency domains.

[0095] The purpose of this step is to further decompose the complex fault traveling wave components into sub-signals with different frequency characteristics, enabling more detailed analysis of the fault signature. Selecting an appropriate modal decomposition algorithm is crucial for improving the accuracy of subsequent fault identification. Depending on the characteristics of the fault traveling wave data, either the empirical mode decomposition (EMD) algorithm or the variational mode decomposition (VMD) algorithm can be used. In specific implementation, the wind farm's weather monitoring module first collects meteorological information, such as wind speed, temperature, and humidity, for the same period as the fault traveling wave data. A weather severity index is then calculated based on this meteorological information. Simultaneously, the unit state complexity is calculated based on the operating status of the wind turbines (such as the number of operating units and power output). Combining weather severity and unit state complexity, a signal smoothness index for the fault traveling wave data is derived. The VMD algorithm is preferred for monitoring collector lines under normal operating conditions. In this case, the signal typically contains stable operating characteristics of multiple wind turbines, and VMD can better separate these characteristics. The EMD algorithm is preferred under severe weather conditions (such as thunderstorms) or when multiple newly connected wind farms or new wind turbine equipment are present. Because strong transient signals may be generated in this case, EMD is more suitable for processing such nonlinear and non-stationary signals.

[0096] The EMD algorithm decomposes the signal into multiple intrinsic mode functions (IMFs) and a residual term through an iterative screening process. Each IMF represents the oscillation mode of the original signal at different time scales. The VMD algorithm, on the other hand, decomposes the signal into a predetermined number of band-limited mode functions through an optimization process. Each mode function has a center frequency and a finite bandwidth. Regardless of the algorithm used, multiple traveling wave modal component signals in different frequency domains are ultimately obtained. These component signals reflect the characteristics of the fault traveling wave in different frequency ranges, providing richer and more detailed information for subsequent feature extraction and fault identification.

[0097] S106. Calculate the signal energy spectra of all traveling wave modal component signals based on the NTEO energy operator, and calibrate the initial wave head of the fault traveling wave data according to the signal energy spectrum of the highest frequency traveling wave modal component signal.

[0098] The NTEO (Normalized Teager Energy Operator) energy operator is an improved energy calculation method that offers superior noise suppression and temporal resolution compared to the traditional Teager energy operator. Each traveling wave modal component signal is first converted into a discrete time series, and then the NTEO energy operator is applied to calculate its energy spectrum. This process generates multiple energy spectra in different frequency domains. Next, the energy spectrum corresponding to the highest-frequency traveling wave modal component signal is selected as the target energy spectrum. Within this energy spectrum, the time point with the highest energy value is identified and calibrated as the initial fault wave head. For example, assuming the sampling frequency of the highest-frequency modal component is 1 MHz, and the energy value at the 1000th sampling point in the energy spectrum is the highest, it can be determined that the initial fault head occurs 1 ms after the fault occurs. The advantage of this method is that the highest-frequency modal component is generally most sensitive to the fault onset time, allowing for more accurate location of the initial fault head. Furthermore, the use of the NTEO energy operator improves location accuracy in noisy environments. Accurately calibrating the initial fault wave head is crucial for subsequent fault location, as it directly impacts the accuracy of fault distance calculation.

[0099] S107. Based on the initial wave head of the fault traveling wave and according to the collector line topology structure, the fault coefficients of all collector branch lines are calculated, and the line fault point in the collector line topology structure is located by combining all the fault coefficients.

[0100] In this approach, the fault data detection terminal node closest to the wind farm transmission system node in the collector line topology is defined as the main line starting node, and the furthest is defined as the main line ending node. The direction from the starting node to the ending node is considered the positive direction of the traveling wave. The propagation time for the line mode component and zero mode component of the fault traveling wave to reach the two terminal nodes is then calculated. Next, based on these distances and the node edge attributes in the topology, the fault coefficient for each collector branch line is calculated. The fault coefficient can be defined as the ratio of the distance from the hypothetical fault point on that branch line to the main line connection point to the total length of the branch line. If a branch line has a fault coefficient of zero, the fault point is determined to be on that branch line, and the midpoint of the branch line is taken as the fault point. If all fault coefficients are non-zero, the fault point is determined to be on the main line, and the two-terminal traveling wave location method is used to determine the specific location. This method combines topological information with traveling wave propagation characteristics to effectively distinguish between main and branch line faults and accurately locate the fault point.

[0101] S108. Extract the energy spectrum features of all signal energy spectra, and identify the fault type of the line fault point based on the energy spectrum features and through a deep learning model.

[0102] First, feature extraction is performed on the energy spectrum of each traveling wave modal component signal. In the time domain, time windows of different sizes are used to intercept energy spectrum segments, from which features such as energy peak, energy decay rate, energy rise time, energy duration, and energy phase difference are extracted. In the frequency domain, statistical features such as energy ratio, energy mean, energy variance, and energy kurtosis are calculated. For example, the energy decay rate can be obtained by fitting the decay curve after the energy peak: E(t) = E0*e (-αt) , where α is the decay rate. Energy phase differences can be quantified by calculating the correlation coefficient of the energy spectra between different phases. Next, correlation analysis is used to filter out the most effective feature subset for fault type identification from these features. This step significantly reduces feature dimensionality, improving model training efficiency and generalization capabilities. Finally, the filtered features are input into a pre-trained deep learning model. This model uses a convolutional neural network architecture with a multi-head attention mechanism to effectively capture the complex relationships between energy features. The model output is the fault type prediction result.

[0103] The selection of the modal decomposition algorithm in step S105 and the improvement of the energy operator in step S106 both significantly improve the accuracy of the model in identifying fault types. By dynamically selecting the empirical mode decomposition (EMD) or variational mode decomposition (VMD) algorithm based on signal stationarity, the signal characteristics under different fault conditions can be better adapted, thereby obtaining more accurate traveling wave modal component signals. These finely decomposed modal components can better reflect the characteristics of the fault signal in different frequency domains, providing a high-quality data foundation for subsequent feature extraction. At the same time, the NTEO (Normalized Teager Energy Operator) energy operator used in S106 has stronger noise suppression capabilities and higher time resolution than the traditional Teager energy operator. This improvement makes the calculation of the energy spectrum more accurate, especially in complex electromagnetic environments, and can more accurately capture the energy variation characteristics of the fault traveling wave. The use of the NTEO energy operator not only improves the positioning accuracy of the initial fault wave head but also provides more reliable energy spectrum data for subsequent feature extraction. The combined optimization of these two key steps significantly enhances the model's ability to distinguish different types of faults, especially when dealing with complex or multiple fault situations, providing more accurate and reliable identification results, thereby significantly improving the performance and reliability of the entire fault monitoring and management system.

[0104] S109. Generate a fault monitoring report based on the line fault point and fault type.

[0105] Among them, the fault location results and fault type identification results obtained in the above steps are integrated to form a comprehensive and detailed fault monitoring report. The content of the report usually includes the following parts: Fault Overview: including basic information such as the time of occurrence, duration, and impact range of the fault. Fault Location: A detailed description of the location of the fault point, including the specific location in the topology (such as the main line or a branch line), the distance to the nearest collector station, etc. Fault Type: Describes the identified fault type (such as single-phase grounding, two-phase short circuit, three-phase short circuit, etc.) and provides the basis for judgment. Fault Feature Analysis: Including technical details such as the main characteristics of the fault traveling wave and energy spectrum characteristics.

[0106] In one embodiment, simplifying the current collection circuit diagram into a current collection circuit topology structure includes the following steps:

[0107] Identify key components in the current collection diagram and the current collection paths between key components;

[0108] All key components are simplified into key nodes, and the collection lines are simplified into key node edges between key nodes to form multiple topological lines. The key nodes include wind turbine nodes, collection station nodes, fault data detection terminal nodes, and wind farm transmission system nodes. The number of fault data detection terminal nodes is 2.

[0109] Add node edge attributes to key node edges based on the actual length of the collector line, and add key node attributes to wind turbine nodes based on the status information of the wind turbines corresponding to the wind turbine nodes.

[0110] For any topological line, if both ends of the topological line are fault data detection terminal nodes, and all other nodes in the topological line except the fault data detection terminal node are collector station nodes, then the topological line is marked as a collector main line, and all collector station nodes are connected to the wind farm transmission system nodes through the collector main line;

[0111] If the two ends of the topological line are the collector station node and the wind turbine generator group node respectively, the topological line is marked as a collector branch line;

[0112] When all topological lines are marked, the simplified collector line topology structure is obtained.

[0113] In this embodiment, the key components typically include wind turbines, collection stations, fault data detection terminals, and wind farm transmission systems. Wind turbines are the main source of electrical energy, collection stations are responsible for collecting and converting electrical energy, fault data detection terminals are used to monitor line status, and the transmission system connects the wind farm to the external power grid. Identifying the collection lines between these components involves tracing the power transmission paths, including underground cables and overhead lines. The key to this step is to accurately record the starting point, end point, and path of each line, laying the foundation for subsequent topology simplification. Graph theory concepts can then be used: each wind turbine, collection station, fault data detection terminal, and transmission system are represented as a node, and the collection lines connecting these nodes are represented as edges.

[0114] Next, you need to add node-edge attributes to key node edges and key node attributes to wind turbine nodes. Node-edge attributes primarily refer to the actual length of the collector line, which is crucial for subsequent fault location calculations. For example, if the line length between two collector stations is 5 kilometers, the attribute value for this edge is set to 5. This length data is typically obtained from engineering drawings or GPS measurements. For wind turbine nodes, key node attributes include the unit's operating status, such as whether it is currently operating, output power, and rotor speed. This information can be obtained in real time from the wind farm's SCADA system. For example, the attributes for a wind turbine node might be: {Grid-connected time: "2020-01-01", Status: "Operating", Power: 2.5MW, Speed: 15rpm}. These attributes not only reflect the real-time operating status of the wind farm but also indicate whether each wind turbine is newly connected to the grid.

[0115] Next, line identification is performed. If a topological line has fault data detection terminal nodes at both ends and only contains a collector station node in the middle, it is marked as a main collector line. This reflects the characteristics of the main power transmission channel in the actual system. Next, the collector branch lines are identified. If a topological line has a collector station node at one end and a wind turbine node at the other end, it is marked as a collector branch line. This represents the power transmission path from a single wind turbine to the collector station. After all lines are marked, the final collector line topology is obtained. This structure clearly shows the power transmission path from the wind turbine to the collector station and then to the transmission system.

[0116] In one embodiment, decomposing the fault traveling wave component using a modal decomposition algorithm to obtain multiple traveling wave modal component signals in different frequency domains includes the following steps:

[0117] The weather monitoring module configured in the wind farm transmission system collects the wind farm meteorological information in the same time period as the fault traveling wave data;

[0118] The weather severity during the period of acquiring fault traveling wave data is calculated based on the wind farm meteorological information and according to preset meteorological condition rules;

[0119] Integrate the key node attributes of all wind turbine nodes, and calculate the unit state complexity of all wind turbine nodes based on all key node attributes;

[0120] The signal stability of the fault traveling wave data is calculated by combining the weather severity and the unit state complexity;

[0121] If the signal stability is less than the preset stability threshold, the empirical mode decomposition algorithm is used to decompose the fault traveling wave component to obtain multiple traveling wave modal component signals in different frequency domains;

[0122] If the signal stability is greater than or equal to the stability threshold, the variational mode decomposition algorithm is used to decompose the fault traveling wave component to obtain multiple traveling wave modal component signals in different frequency domains.

[0123] In this embodiment, collecting wind farm meteorological information in the same time period as the fault traveling wave data through a weather monitoring module configured in the wind farm transmission system is a key step. The weather monitoring module usually includes devices such as anemometers, thermometers, humidity sensors, and barometers, which are distributed in different locations of the wind farm to fully capture local meteorological changes. When the fault traveling wave data is detected, the system will synchronously record the current meteorological data. For example, if the fault occurs at 14:30:25 on July 15, 2023, then the wind speed (such as 12m / s), temperature (such as 28°C), humidity (such as 65%) and air pressure (such as 101.3kPa) at this precise moment will be recorded. These real-time meteorological data are crucial for subsequent analysis of fault causes and signal characteristics, because different meteorological conditions may have a significant impact on the operating status and fault characteristics of the power system. In this way, the association between fault phenomena and environmental factors can be established, providing an important basis for more accurate fault diagnosis and prevention.

[0124] Based on the collected wind farm meteorological information, the weather severity during the fault traveling wave data acquisition period is calculated according to preset meteorological condition rules. This step aims to quantify the potential impact of meteorological conditions on power system operation. The calculation of weather severity usually involves a weighted combination of multiple factors. For example, the following formula can be used:

[0125] Weather severity = a*(wind speed / wind speed threshold) + b*|temperature-suitable temperature| / temperature range + c*(humidity / humidity threshold) + d*|air pressure-standard air pressure| / pressure range.

[0126] Among them, a, b, c, and d are weight coefficients, determined based on the importance of each factor to the system. Parameters such as wind speed threshold, optimum temperature, humidity threshold, and standard atmospheric pressure are determined based on equipment specifications and historical data. For example, if the wind speed is 25 m / s (exceeding the 20 m / s threshold), the temperature is 35°C (deviating from the optimum temperature of 25°C), the humidity is 90% (exceeding the 80% threshold), and the air pressure is 98 kPa (deviating from the standard atmospheric pressure of 101.3 kPa), the calculated weather severity may be higher, such as 0.8 (assuming a full score of 1). This indicator intuitively reflects the severity of the current weather conditions and provides an important reference for subsequent signal processing and fault analysis.

[0127] Next, the key node attributes of all wind turbine nodes are integrated and the unit state complexity of all wind turbine nodes is calculated based on these attributes. This step aims to quantify the complexity of the overall operating state of the wind farm. Key node attributes usually include the operating state and grid connection time of each wind turbine. The operating state can be divided into several categories, such as normal operation, fault shutdown, maintenance, etc. The grid connection time reflects the degree of impact of the new wind turbine on the fluctuation of the wind farm line. Therefore, the unit state complexity can be calculated using the following method:

[0128] Unit state complexity = Σ(Wi*Si*Ti) / N

[0129] Where Wi is the weight of each operating state, Si is the number of wind turbines in that state, Ti is the average grid-connected time factor for the group of wind turbines, and N is the total number of wind turbines. This metric reflects the overall complexity of the wind farm's current operating state and provides important background information for subsequent signal analysis.

[0130] Next, the signal smoothness of the fault traveling wave data is calculated based on the weather severity and unit state complexity. The purpose of this step is to evaluate the overall characteristics of the fault traveling wave data and provide a basis for selecting appropriate signal processing methods. The signal smoothness can be calculated using the following formula:

[0131] Signal stability = 1-(α*weather severity+β*unit state complexity) / (α+β)

[0132] Here, α and β are weighting coefficients used to adjust the relative impact of meteorological factors and unit status on signal characteristics. For example, if the weather severity is 0.8 and the unit state complexity is 1.11, α = 0.6 and β = 0.4, then the signal smoothness is: 1-(0.6*0.8+0.4*1.11) / (0.6+0.4) = 0.0734. The closer this value is to 1, the smoother the signal; the closer it is to 0, the less stable the signal. Based on the calculated signal smoothness, an appropriate modal decomposition algorithm is selected to process the fault traveling wave component. If the signal smoothness is less than a preset smoothness threshold (e.g., 0.5), the empirical mode decomposition (EMD) algorithm is used; if it is greater than or equal to the threshold, the variational mode decomposition (VMD) algorithm is used. Both algorithms aim to decompose the fault signal into multiple single-frequency components, but are suitable for different signal characteristics. The VMD algorithm is preferred for monitoring collector lines under normal operating conditions. Because the signal in this case often contains the steady-state operating characteristics of multiple wind turbines, VMD is better able to separate these features. EMD is the preferred algorithm in severe weather conditions (such as thunderstorms) and when there are multiple newly connected wind farms or new wind turbines. Because these situations can produce strong transient signals, EMD is more suitable for handling such nonlinear and non-stationary signals.

[0133] The EMD algorithm decomposes a signal into several intrinsic mode functions (IMFs) and a residual term through an iterative "screening" process. It is particularly well-suited for processing nonlinear and non-stationary signals. For example, for a fault traveling wave signal containing multiple frequency components, EMD may decompose it into 5-10 IMFs, each representing an oscillation mode within a specific frequency range. The VMD algorithm, on the other hand, decomposes the signal into a predetermined number of band-limited mode functions through an optimization process. It may be more stable than EMD for certain signal types, particularly when processing near-stationary signals. For example, VMD may decompose the same fault traveling wave signal into 3-5 mode functions, each with a well-defined center frequency and bandwidth. Regardless of the algorithm used, multiple traveling wave modal component signals in different frequency domains are ultimately obtained. These component signals reflect the characteristics of the fault traveling wave in different frequency ranges, providing a more refined and reliable information basis for subsequent feature extraction and fault identification. This adaptive algorithm selection allows for more accurate processing of complex fault traveling wave data, improving the accuracy and reliability of subsequent analysis.

[0134] In one embodiment, decomposing the fault traveling wave component using an empirical mode decomposition algorithm to obtain multiple traveling wave modal component signals in different frequency domains includes the following steps:

[0135] The signal maximum and minimum of the fault traveling wave component are calculated, and the upper and lower envelopes of the fault traveling wave component are obtained by combining the signal maximum and signal minimum and using the cubic spline interpolation method.

[0136] The upper and lower envelopes are averaged to obtain the average component of the fault traveling wave component;

[0137] The difference between the fault traveling wave component and the average component is used as the preliminary IMF component;

[0138] If the number of zero points and the number of extreme values ​​of the prepared IMF component are equal, or the average values ​​of the upper and lower envelopes of the prepared IMF component at any time are both 0, then the prepared IMF component is regarded as one of the traveling wave mode component signals in the frequency domain;

[0139] Calculate the remaining signal components after removing the traveling wave modal component signal from the fault traveling wave component;

[0140] The remaining signal components are used as the fault traveling wave components of a new round and the above component decomposition steps are repeated until the component decomposition cutoff condition is met;

[0141] After completing the component decomposition of the fault traveling wave component, the soft limiting function is used to filter out the signal noise of all traveling wave modal component signals in the corresponding frequency domain.

[0142] In this embodiment, the signal maximum and minimum of the fault traveling wave component are first calculated, and the upper and lower envelopes are obtained using cubic spline interpolation. Specifically, the fault traveling wave component data is traversed to find all local maximum and minimum points. For example, for a signal containing 1000 sampling points, 50 maximum points and 49 minimum points may be found. Then, these extreme points are connected using the cubic spline interpolation method to form the upper and lower envelopes. Cubic spline interpolation can generate a smooth curve, avoiding the jagged effect that may be caused by linear interpolation. The upper and lower envelopes are then averaged to obtain the average component of the fault traveling wave component. The specific operation is to add the values ​​of the upper envelope and the lower envelope at each time point and divide by 2. For example, if the value of the upper envelope is 10 and the value of the lower envelope is -8 at a certain moment, the average component value at that moment is 1. This average component reflects the overall trend of the signal. Removing this trend can highlight the oscillation component in the signal.

[0143] Next, the difference between the fault traveling wave component and the average component is used as the preliminary IMF component. This step extracts the oscillatory components in the signal. Specifically, the average component just calculated is subtracted from the original fault traveling wave component. For example, if the original signal value is 5 and the average component value is 1 at a certain moment, the value of the preliminary IMF component at that moment is 4. This difference represents the local oscillation in the signal after removing the overall trend and is a potential intrinsic mode function (IMF). Therefore, it is necessary to determine whether the preliminary IMF component meets the definition criteria of an IMF. An IMF must meet two conditions: the number of extreme points and the number of zero points are equal or differ by no more than 1; and the average values ​​of the upper and lower envelopes are close to 0 at all times. If these conditions are met, the preliminary IMF component is determined to be a frequency-domain traveling wave modal component signal. For example, if the preliminary IMF component has 100 extreme points and 99 zero points, and the average values ​​of its upper and lower envelopes are less than a preset threshold (such as 0.01) at all times, it can be considered a valid IMF.

[0144] Next, the remaining signal components after removing the traveling wave modal component signals from the fault traveling wave components are calculated. This is done by subtracting the newly determined IMF from the original fault traveling wave components. This remaining component contains low-frequency components and trends that have not yet been extracted. The remaining signal components are used as the new round of fault traveling wave components and the above component decomposition steps are repeated until the component decomposition cutoff condition is met. This process is the iterative core of the EMD algorithm. The component decomposition cutoff condition is expressed as follows:

[0145]

[0146] Where: M represents the number of decomposed traveling wave modal component signals, represents the traveling wave modal component signal obtained in the mth round of component decomposition step, represents the traveling wave modal component signal obtained in the m-1th round of the component decomposition step, and ε represents the cutoff threshold. This iterative process ensures that all frequency components in the signal are extracted one by one.

[0147] After completing the component decomposition of the fault traveling wave components, the soft limiting function is used to filter out the signal noise of all traveling wave modal component signals in the corresponding frequency domain. This step is intended to improve the signal-to-noise ratio of the signal. The soft limiting function can be defined as:

[0148] f(x)=sign(x)*max(0,|x|-threshold)

[0149] The threshold is set based on the characteristics of each IMF. For example, for high-frequency IMFs, a higher threshold can be set to remove more noise. This step effectively removes random noise in each frequency band while preserving the signal's key characteristics, thereby improving the accuracy of subsequent analysis.

[0150] In one embodiment, decomposing the fault traveling wave component using a variational mode decomposition algorithm to obtain multiple traveling wave modal component signals in different frequency domains includes the following steps:

[0151] A decomposition constraint model of the fault traveling wave component is constructed based on the decomposition constraints of the variational mode decomposition algorithm. The decomposition constraints are as follows: the N traveling wave modal component signals obtained by decomposing the fault traveling wave component using the variational mode decomposition algorithm must have the smallest deviation from the fault traveling wave component after reconstruction, and the sum of the estimated bandwidths of all traveling wave modal component signals must be the smallest.

[0152] The Lagrangian operator and penalty factor are introduced into the decomposition constraint model to construct the model augmentation expression of the decomposition constraint model;

[0153] Based on the model augmentation expression and the alternating direction method of the Lagrangian operator, alternating iterative optimization is performed in the frequency domain until the Wiener filtering of all traveling wave modal components meets the convergence conditions. Each round of alternating iterative optimization process decomposes a traveling wave modal component signal in a different frequency domain from the fault traveling wave component.

[0154] In this embodiment, a decomposition constraint model for the fault traveling wave components is constructed based on the decomposition constraints of the variational modal decomposition algorithm, thereby decomposing the fault traveling wave components in an optimized manner. The decomposition constraints include two key objectives: first, the reconstructed N decomposed traveling wave modal component signals should have the minimum deviation from the original fault traveling wave components to ensure the accuracy of the decomposition; second, the sum of the estimated bandwidths of all traveling wave modal component signals should be minimized, which ensures the simplicity of the decomposition. Specifically, it can be expressed as the minimization objective function:

[0155]

[0156] Among them, uk(t) is the kth modal component, ωk is its center frequency, and f(t) is the original signal. This model ensures both reconstruction accuracy and minimization of the bandwidth of each component by optimizing simultaneously in the time and frequency domains, thereby achieving accurate decomposition of complex fault traveling wave signals. Then, the Lagrangian operator and penalty factor are introduced into the decomposition constraint model to construct the model augmentation expression of the decomposition constraint model. The purpose of this step is to transform the constrained optimization problem into an unconstrained optimization problem for easy solution. Specifically, the Lagrangian multiplier λ and the penalty factor α are introduced to construct the augmented Lagrangian function:

[0157]

[0158] Here, α controls the strength of the bandwidth constraint, and the λ / 2 term ensures that the reconstruction constraint is satisfied. This formulation integrates the original constraints into the objective function, allowing the optimization process to consider all constraints simultaneously, greatly simplifying the solution. By adjusting α and λ, a balance can be achieved between reconstruction accuracy and bandwidth minimization, resulting in the optimal decomposition result.

[0159] Then, based on the model augmentation expression and using the alternating direction method of the Lagrangian operator, alternating iterative optimization is performed in the frequency domain until the Wiener filter of all traveling wave modal components meets the convergence conditions. Each round of alternating iterative optimization process decomposes the fault traveling wave component into a traveling wave modal component signal in a different frequency domain. This process is the core of the variational mode decomposition algorithm, and iterative optimization is used to gradually extract each frequency component. The specific process is as follows:

[0160] a) Initialize each modal component uk and center frequency ωk.

[0161] b) For each k, update uk: uk = argminL(ui, ωi, λ), i≠k fixed.

[0162] c) Update ωk: ωk = argminL(ui, ωi, λ), i≠k fixed.

[0163] d) Update λ: λ=λ+τ(f-∑uk).

[0164] e) Repeat bd until convergence.

[0165] In each iteration, uk is updated by the Wiener filter.

[0166] The convergence condition is expressed as follows:

[0167]

[0168] Where: K represents the number of traveling wave modal component signals that have been decomposed, represents the Wiener filtering of the traveling wave modal component signal obtained during the s-th round of alternating iterative optimization, represents the Wiener filtering of the traveling wave modal component signal obtained in the s+1th round of alternating iterative optimization, ||·||2 represents the Euclidean norm, and ζ represents the convergence accuracy.

[0169] This process ensures that each modal component is an optimal representation of a specific frequency band in the original signal. Through continuous iteration, the algorithm gradually separates the individual frequency components of the fault traveling wave signal until all components meet convergence criteria. This method effectively handles nonlinear and nonstationary signals and is particularly suitable for analyzing complex fault traveling wave data.

[0170] In one embodiment, calculating the signal energy spectra of all traveling wave modal component signals based on the NTEO energy operator and calibrating the fault traveling wave initial wave head of the fault traveling wave data according to the signal energy spectrum of the highest frequency traveling wave modal component signal includes the following steps:

[0171] Convert all traveling wave modal component signals into discrete component signals;

[0172] The signal energy spectrum of each discrete component signal is calculated using the NTEO energy operator;

[0173] The signal energy spectrum corresponding to the traveling wave modal component signal in the highest frequency domain is selected as the target signal energy spectrum, and the moment with the highest energy value in the target signal energy spectrum is calibrated as the initial wave head of the fault traveling wave data.

[0174] In this embodiment, all traveling wave modal component signals are converted into discrete component signals. This process is actually to convert a continuous analog signal into a series of discrete digital representations. In specific implementation, the appropriate sampling frequency is first determined, which should be at least twice the highest frequency component of the signal (following the Nyquist sampling theorem) to avoid information loss and aliasing effects. For example, if the highest frequency component is 5kHz, the sampling frequency should be at least 10kHz. Then, the continuous signal is sampled at equally spaced time points. Assuming there is a traveling wave modal component signal lasting 1 second and the sampling frequency is 10kHz, then 10,000 discrete data points will be obtained. This conversion makes subsequent digital signal processing possible while retaining the key features of the original signal. The converted discrete signal can be expressed as h(x), where x is the discrete time index.

[0175] Next, we use the NTEO (Normalized Teager Energy Operator) energy operator to calculate the signal energy spectrum of each discrete component signal. NTEO is an improved version of the Teager energy operator, with stronger noise immunity and higher time resolution. For each discrete component signal, the formula for calculating the signal energy spectrum using NTEO is as follows:

[0176]

[0177] Where: Ψ[h i (x)] represents the i-th discrete component signal h i (x) represents the signal energy spectrum, where T represents the resolution parameter and M represents the number of discrete component signals. This operator effectively captures instantaneous energy changes in the signal. These energy values ​​constitute the signal's energy spectrum, intuitively reflecting the signal's energy distribution at different moments. NTEO is particularly well-suited for processing nonlinear and non-stationary signals because it can accurately locate sudden and transient changes in the signal. This method generates a corresponding energy spectrum for each discrete component signal, providing an important basis for subsequent fault location.

[0178] Next, the signal energy spectrum corresponding to the traveling wave modal component signal in the highest frequency domain is selected as the target signal energy spectrum, and the point in time with the highest energy value in the target signal energy spectrum is calibrated as the initial wave head of the fault traveling wave data. The core idea of ​​this step is to accurately locate the moment of fault occurrence by leveraging the sensitivity of high-frequency components to sudden changes in the system. The highest-frequency components are selected because high-frequency signals typically contain the richest transient information and are most sensitive to sudden changes in the system. The point with the highest energy value in the selected energy spectrum is considered the initial wave head of the fault traveling wave, as a sudden energy surge typically occurs when a fault occurs. Assuming that the energy value of the 500th sampling point in the energy spectrum is the highest and the sampling frequency is 10kHz, it can be determined that the fault occurred 50ms after the signal onset. The advantage of this method is that it can very accurately locate the fault moment, with an error typically within milliseconds, providing an accurate time reference for subsequent fault type identification and fault location.

[0179] In one embodiment, the fault traveling wave component includes a fault traveling wave line mode component and a fault traveling wave zero mode component. The fault traveling wave line mode component and the fault traveling wave zero mode component share an initial wave head of the fault traveling wave. Based on the initial wave head of the fault traveling wave and according to the collector line topology, the fault coefficients of all collector branch lines are calculated. Locating the line fault point in the collector line topology by combining all the fault coefficients includes the following steps:

[0180] The fault data detection terminal node closest to the wind farm transmission system node in the collector line topology is used as the main line starting node, and the fault data detection terminal node farthest from the wind farm transmission system node is used as the main line ending node;

[0181] The direction from the main line starting node to the main line ending node is defined as the positive direction of the traveling wave;

[0182] Combined with the initial wave head of the fault traveling wave and the first propagation velocity of the fault traveling wave line mode component, the first propagation time of the fault traveling wave line mode component to the starting node of the main line and the second propagation time of the fault traveling wave line mode component to the ending node of the main line are calculated respectively;

[0183] Combined with the initial wave head of the fault traveling wave and the second propagation velocity of the zero-mode component of the fault traveling wave, the third propagation time of the zero-mode component of the fault traveling wave to the starting node of the main line and the fourth propagation time of the zero-mode component of the fault traveling wave to the ending node of the main line are calculated respectively;

[0184] By combining the first propagation time, the second propagation time, the third propagation time, the fourth propagation time and the first propagation speed, a first fault distance between the hypothetical fault point and the starting node of the main line and a second fault distance between the hypothetical fault point and the ending node of the main line are calculated respectively;

[0185] Combining the first fault distance, the second fault distance and the node edge attributes in the collector line topology structure, and calculating the fault coefficient of all collector branch lines according to the positive direction of the traveling wave;

[0186] If the fault coefficient of any collector branch line is 0, the line fault point generating the fault traveling wave data is determined to be located on the fault collector branch line with a fault coefficient of 0, and the midpoint of the fault collector branch line is taken as the line fault point in the collector line topology structure;

[0187] If all fault coefficients are not zero, the line fault point is determined to be located on the collector main line. The line fault point in the collector line topology is determined by combining the initial wave head of the fault traveling wave, the propagation speed of the fault traveling wave component, and the node edge attribute calculation, and using the double-terminal traveling wave location method.

[0188] In this embodiment, due to the characteristics of wind farm collection lines with many branches and short lines, it is difficult for traditional single-ended and double-ended methods to accurately calculate the fault distance, resulting in large errors in the calculation results and misjudgment of the fault branch. Therefore, this embodiment adopts an improved single-ended traveling wave method to calculate the fault coefficient of the line. Figure 2, the fault data detection terminal node closest to the wind farm transmission system node in the collector line topology is used as the main line starting node Q1, and the fault data detection terminal node farthest from the wind farm transmission system node is used as the main line ending node Q n , determine the main collector line Q1Qn, and then define the direction from the main line starting node to the main line ending node as the traveling wave positive direction. Starting from the main line starting node Q1, along the traveling wave positive direction and in the order of small to large, the collector station nodes on the collector main line are marked as the branch line starting nodes {p1, p2, p3, ..., p N}, N represents the number of nodes in the collector station. The end of each collector branch line is the last wind turbine node in the collector branch line. The end nodes of the collector branch line are marked as {q1, q2, q3, ..., q N}.

[0189] Then, the initial time t0 of the fault occurrence is determined based on the initial wave head of the fault traveling wave. Then, the initial time t0 of the fault occurrence and the first propagation velocity of the fault traveling wave line mode component are combined to respectively calculate the first propagation time of the fault traveling wave line mode component to propagate to the starting node of the main line (the propagation time refers to the signal propagation time when the signal propagates from the initial time of the fault occurrence through the line to the corresponding node. The propagation times described subsequently in this embodiment have the same meaning), and the second propagation time of the fault traveling wave line mode component to propagate to the main line terminal node. Next, the initial time t0 of the fault occurrence and the second propagation velocity of the fault traveling wave zero mode component are combined to respectively calculate the third propagation time of the fault traveling wave zero mode component to propagate to the starting node of the main line, and the fourth propagation time of the fault traveling wave zero mode component to propagate to the main line terminal node. Then, based on the above calculation results, the first fault distance between the hypothetical fault point and the starting node of the main line can be further calculated. The hypothetical fault point is the fault occurrence point to be specifically located in the collector line. The calculation formula for the first fault distance is as follows:

[0190]

[0191] Where: represents the first fault distance between the hypothetical fault point G and the main line starting node Q1, t1 represents the first propagation time, t2 represents the second propagation time, t3 represents the third propagation time, t4 represents the fourth propagation time, and v1 represents the first propagation speed.

[0192] The second fault distance between the hypothetical fault point and the main line termination node is further calculated. The calculation formula for the second fault distance is as follows:

[0193]

[0194] Where: Indicates the hypothetical fault point G and the main line termination node Q n The second fault distance between

[0195] like Figure 2 As shown, the main collector line can be considered a special branch consisting of branches Q1p1, p1p2, p2p3, p3p4, p4p5, and p5Qn. After a fault occurs, because the node edge attributes in the collector line topology represent the actual line length of each branch, the fault coefficient of each wind turbine branch can be calculated by combining the first fault distance, the second fault distance, and the node edge attributes in the collector line topology. If the fault occurs in a collector branch line, the fault coefficients of each collector branch line calculated using the above method are shown in Table 1 below.

[0196] Table 1 Failure coefficients of each collector branch line when a collector branch line fails

[0197]

[0198] According to the results in Table 1, the fault coefficient of the faulty collector branch line is 0, while the fault coefficient of the unfaulted collector branch line is not 0. Taking the faulty collector branch line as the reference point, the decision coefficient of the collector branch line near the main line starting node along the positive direction of the traveling wave is greater than 0, while the decision coefficient of the collector branch line near the main line ending node is less than 0. Therefore, it can be further confirmed that if the decision coefficient of a collector branch line is 0, and the decision coefficients of the remaining collector branch lines are not 0, it means that the fault occurred in the collector branch line with a fault coefficient of 0. Because the collector branch line is short, the midpoint of the collector branch line can be simply used as the final line fault location point. If the fault occurs in the main collector line, the fault coefficients of each collector branch line calculated using the above content are shown in Table 2 below.

[0199] Table 2 Failure coefficients of each collector branch line when the collector main line fails

[0200]

[0201] According to the results in Table 2, when a fault occurs on the main collector line, the decision coefficients of all collector branch lines are all 0, and the value of the decision coefficient is related to the location of the hypothetical fault point. The decision coefficient of the collector branch line near the starting node of the main line along the positive direction of the traveling wave is greater than 0, and the decision coefficient of the collector branch line near the ending node of the main line is less than 0. Therefore, it can be further confirmed that if all fault coefficients are not 0, the line fault point is located on the main collector line. Because the main collector line is very long and the calculated fault distance may correspond to multiple fault locations, it is necessary to further use a more accurate dual-end traveling wave positioning method for precise positioning. The calculation formula of the dual-end traveling wave positioning method used in this embodiment is as follows:

[0202]

[0203] Where: Indicates the actual length of the collector main line, It represents the propagation time of the fault traveling wave generated by the line fault point to the starting node of the main line. It represents the propagation time of the fault traveling wave generated by the line fault point to the terminal node of the main line, and v represents the propagation speed of the fault traveling wave generated by the line fault point.

[0204] In one embodiment, extracting energy spectrum features of all signal energy spectra and identifying the fault type of the line fault point based on the energy spectrum features using a deep learning model includes the following steps:

[0205] For each traveling wave modal component signal's signal energy spectrum, energy spectrum segments of different time lengths are cut out from the signal energy spectrum through time windows of different time scales, and energy time domain features are extracted from each energy spectrum segment. The energy time domain features include energy peak value, energy decay rate, energy rise time, energy duration, and energy phase difference features.

[0206] Statistical methods are used to extract the energy frequency domain features of the signal energy spectrum in the frequency domain. The energy frequency domain features include energy ratio, energy mean, energy variance and energy kurtosis.

[0207] Use correlation analysis to select energy feature subsets with fault type discrimination capabilities from energy time domain features and energy frequency domain features;

[0208] The energy feature subset is input into the pre-trained energy feature recognition model, and the fault type of the line fault point is output through the energy feature recognition model. The energy feature recognition model is a convolutional neural network model that includes a multi-head attention mechanism.

[0209] In this embodiment, multi-scale time window analysis of the energy spectrum of each traveling wave modal component signal is a key step in extracting rich feature information. This process involves using time windows of different lengths (such as 10ms, 50ms, 100ms, etc.) to slide through the entire energy spectrum, thereby capturing the energy change characteristics at different time scales. For each energy spectrum segment intercepted by the window, five energy time domain features are extracted: (1) Energy peak, that is, the maximum energy value in the segment, reflecting the fault intensity; (2) Energy decay rate, the energy decline trend is calculated by exponential fitting, characterizing the fault persistence; (3) Energy rise time, the time from the energy start to increase to the peak, indicating the fault development speed; (4) Energy duration, the duration of energy exceeding a certain threshold, reflecting the scope of fault impact; (5) Energy phase difference characteristics, comparing the energy differences between different phases, which helps to identify fault types such as phase short circuit. These features comprehensively characterize the energy distribution characteristics of the fault traveling wave in the time domain, providing an important basis for subsequent fault type discrimination.

[0210] Next, using statistical methods to extract the frequency domain features of the signal energy spectrum is another important dimension for analyzing fault characteristics. This step focuses on the statistical characteristics of energy in frequency distribution, including four key indicators: (1) Energy ratio, which calculates the ratio of energy in a specific frequency band to the total energy, reflecting the frequency concentration of the fault signal; (2) Energy mean, which calculates the average energy level of the entire spectrum, characterizing the overall strength of the fault; (3) Energy variance, which measures the discreteness of the energy distribution and indicates the frequency complexity of the fault signal; (4) Energy kurtosis, which describes the sharpness of the energy distribution and helps to identify sudden faults. The calculation of these features usually requires first performing a Fourier transform on the energy spectrum, converting the energy spectrum to the frequency domain, and then applying the corresponding statistical formula to calculate them. These frequency domain features, combined with time domain features, provide a comprehensive description of the fault signal, enhancing the accuracy and reliability of subsequent fault type identification.

[0211] Next, using correlation analysis to select a subset of features with fault type discriminative power from the energy time-domain and frequency-domain features is a key step in optimizing the feature space. This process aims to reduce redundant information and improve the efficiency and accuracy of fault identification. In specific implementation, methods such as the Pearson correlation coefficient or mutual information can be used to calculate the correlation between each feature and the fault type, as well as the mutual correlation between features. For example, the correlation coefficient between each feature and the fault type is calculated, and features with an absolute value of the correlation coefficient greater than 0.5 are selected. At the same time, for feature pairs with a mutual correlation coefficient greater than 0.8, only the one with the higher correlation with the fault type is retained. This step not only improves the training efficiency of subsequent models, but also reduces the risk of overfitting and enhances the model's generalization ability on new data. The selected feature subset includes energy peaks, energy ratios of major frequency bands, and energy decay rates. These features together constitute a concise and effective fault feature representation.

[0212] Finally, the selected energy feature subset is input into a pre-trained energy feature recognition model to output the fault type of the line fault point. This step utilizes a convolutional neural network model with a multi-head attention mechanism. This architecture combines the advantages of CNNs in extracting local features with the ability of the attention mechanism to capture long-range dependencies. The model's structure is as follows: First, the input layer receives the feature subset. Next, high-level features are extracted through multiple convolutional and pooling layers. Specifically, there are three convolutional layers, each using kernels of different sizes to capture feature patterns at different scales. Next, a multi-head attention layer calculates the correlation between different features. Using three attention heads, each head independently learns the associations between features. Finally, a fully connected layer and a softmax layer output a probability distribution of the fault type. This model architecture effectively utilizes the extracted energy features to automatically learn the complex patterns of fault types, thereby achieving high-accuracy fault type recognition. The model training process involves a large amount of labeled historical fault data. The network parameters are continuously adjusted through a backpropagation algorithm to ultimately achieve optimal performance on the validation set. Through the above process, starting from the original traveling wave modal component signal, through multi-scale time domain analysis, frequency domain statistical feature extraction, correlation-based feature selection, and finally using advanced deep learning models to identify fault types, a comprehensive, efficient, and accurate fault diagnosis system has been constructed. This system can quickly process complex fault traveling wave data, extract key features, and provide reliable fault type judgments.

[0213] The present invention also discloses a fault monitoring and management system for a 35kV wind farm overhead collector line, comprising a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the computer program, it implements the fault monitoring and management method for the 35kV wind farm overhead collector line as described in any one of the above embodiments.

[0214] Among them, the processor can adopt a central processing unit (CPU). Of course, according to actual usage, other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. can also be adopted. The general-purpose processor can adopt a microprocessor or any conventional processor, etc., and this application does not impose any restrictions on this.

[0215] Among them, the memory can be an internal storage unit of a computer device, such as a hard disk or memory of a computer device, or an external storage device of a computer device, such as a plug-in hard disk, smart memory card (SMC), secure digital card (SD) or flash memory card (FC) equipped on the computer device. In addition, the memory can also be a combination of an internal storage unit and an external storage device of a computer device. The memory is used to store computer programs and other programs and data required by the computer device. The memory can also be used to temporarily store data that has been output or is to be output. This application does not impose any restrictions on this.

[0216] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of protection of the present application is limited to these examples. In line with the present application, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of different aspects of one or more embodiments of the present application as above, which are not provided in detail for the sake of simplicity.

[0217] The one or more embodiments of this application are intended to encompass all such substitutions, modifications, and variations that fall within the broad scope of this application. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of one or more embodiments of this application should be included in the scope of protection of this application.

Claims

1. A fault monitoring and management method for a 35kV wind farm overhead collector line, characterized in that: The steps include: Obtaining a collection line diagram of a wind farm collection line; Simplifying the collector line diagram into a collector line topology structure, wherein the collector line topology structure includes a collector main line directly connected to the wind farm power transmission system, and a plurality of collector branch lines connected to the collector main line; Acquiring fault traveling wave data in the wind farm collector line through a fault data detection terminal deployed in the collector main line; Decoupling the fault traveling wave data into fault traveling wave components using a Karen Boolean transform method; Decomposing the fault traveling wave component by using a modal decomposition algorithm to obtain a plurality of traveling wave modal component signals in different frequency domains; The signal energy spectrum of all the traveling wave modal component signals is calculated based on the NTEO energy operator, and the initial wave head of the fault traveling wave of the fault traveling wave data is calibrated according to the signal energy spectrum of the highest frequency traveling wave modal component signal; Based on the initial wave head of the fault traveling wave and according to the collector line topology structure, the fault coefficients of all the collector branch lines are calculated, and the line fault point in the collector line topology structure is located by combining all the fault coefficients; Extracting energy spectrum features of all the signal energy spectra, and identifying the fault type of the line fault point based on the energy spectrum features and using a deep learning model; A fault monitoring report is generated in combination with the line fault point and the fault type.

2. The fault monitoring and management method for a 35kV wind farm overhead collector line according to claim 1, characterized in that: Simplifying the current collecting circuit diagram into a current collecting circuit topology structure comprises the following steps: Identify key components in the current collection circuit diagram and current collection circuits between the key components; Simplifying all the key components into key nodes, and simplifying the collection lines into key node edges between the key nodes to form multiple topological lines, wherein the key nodes include wind turbine nodes, collection station nodes, fault data detection terminal nodes, and wind farm transmission system nodes, and the number of the fault data detection terminal nodes is 2; adding node edge attributes to the key node edges according to the actual length of the collector line, and adding key node attributes to the wind turbine generator set nodes according to the unit status information of the wind turbine generator set corresponding to the wind turbine generator set nodes; For any of the topological lines, if both ends of the topological line are the fault data detection terminal nodes, and all the nodes in the topological line except the fault data detection terminal node are the collector station nodes, then the topological line is marked as a collector main line, and all the collector station nodes are connected to the wind farm transmission system nodes through the collector main line; If the two ends of the topological line are the collector station node and the wind turbine generator set node respectively, the topological line is marked as a collector branch line; When all the topological lines are marked, the simplified collector line topological structure of the collector line diagram is obtained.

3. The fault monitoring and management method for a 35kV wind farm overhead collector line according to claim 2, characterized in that: Decomposing the fault traveling wave component by using a modal decomposition algorithm to obtain a plurality of traveling wave modal component signals in different frequency domains comprises the following steps: Collecting wind farm meteorological information of the same time period as the fault traveling wave data by a weather monitoring module configured in the wind farm transmission system; Calculating the weather severity during the period of acquiring the fault traveling wave data based on the wind farm weather information and according to preset weather condition rules; Integrating the key node attributes of all the wind turbine generator set nodes, and calculating the unit state complexity of all the wind turbine generator set nodes according to all the key node attributes; Calculating the signal stability of the fault traveling wave data by combining the weather severity and the unit state complexity; If the signal stability is less than a preset stability threshold, an empirical mode decomposition algorithm is used to decompose the fault traveling wave component to obtain a plurality of traveling wave modal component signals in different frequency domains; If the signal stability is greater than or equal to the stability threshold, a variational mode decomposition algorithm is used to decompose the fault traveling wave component to obtain a plurality of traveling wave modal component signals in different frequency domains.

4. The fault monitoring and management method for a 35kV wind farm overhead collector line according to claim 3, characterized in that: Decomposing the fault traveling wave component by using the empirical mode decomposition algorithm to obtain a plurality of traveling wave modal component signals in different frequency domains includes the following steps: Calculating the signal maximum and signal minimum of the fault traveling wave component, combining the signal maximum and signal minimum and using a cubic spline interpolation method to obtain the upper envelope and lower envelope of the fault traveling wave component; Performing an average calculation on the upper envelope and the lower envelope to obtain an average component of the fault traveling wave component; taking the difference between the fault traveling wave component and the average component as a preliminary IMF component; If the number of zero points and the number of extreme values ​​of the prepared IMF component are equal, or the average values ​​of the upper and lower envelopes of the prepared IMF component at any time are both 0, then the prepared IMF component is used as a traveling wave modal component signal in one of the frequency domains; Calculating the remaining signal component after subtracting the traveling wave modal component signal from the fault traveling wave component; The remaining signal components are used as a new round of fault traveling wave components and the above component decomposition steps are repeated until a component decomposition cutoff condition is met; After completing the component decomposition of the fault traveling wave component, a soft limit function is used to filter out the signal noise of all the traveling wave modal component signals in the corresponding frequency domain; The component decomposition cutoff condition is expressed as follows: Where: M represents the number of decomposed traveling wave modal component signals, represents the traveling wave modal component signal obtained in the mth round of component decomposition step, represents the traveling wave modal component signal obtained in the m-1th round of component decomposition step, and ε represents the cutoff threshold.

5. The fault monitoring and management method for a 35kV wind farm overhead collector line according to claim 3, characterized in that: Decomposing the fault traveling wave component by using a variational modal decomposition algorithm to obtain a plurality of traveling wave modal component signals in different frequency domains comprises the following steps: A decomposition constraint model of the fault traveling wave component is constructed based on the decomposition constraint conditions of the variational modal decomposition algorithm, wherein the decomposition constraint conditions are as follows: N traveling wave modal component signals obtained by decomposing the fault traveling wave component using the variational modal decomposition algorithm have the smallest deviation from the fault traveling wave component after reconstruction, and the sum of the estimated bandwidths of all the traveling wave modal component signals is the smallest; Introducing a Lagrangian operator and a penalty factor into the decomposition constraint model to construct a model augmented expression of the decomposition constraint model; Based on the model augmented expression and using the alternating direction method of the Lagrangian operator, alternating iterative optimization is performed in the frequency domain until the Wiener filtering of all the traveling wave modal components meets the convergence condition, and each round of alternating iterative optimization process decomposes the traveling wave modal component signal of a different frequency domain from the fault traveling wave component; The convergence condition is expressed as follows: Where: K represents the number of traveling wave modal component signals that have been decomposed, represents the Wiener filtering of the traveling wave modal component signal obtained during the s-th round of alternating iterative optimization, represents the Wiener filtering of the traveling wave modal component signal obtained in the s+1th round of alternating iterative optimization, ||·||2 represents the Euclidean norm, and ζ represents the convergence accuracy.

6. The fault monitoring and management method for a 35kV wind farm overhead collector line according to claim 1, characterized in that: The steps of calculating the signal energy spectra of all the traveling wave modal component signals based on the NTEO energy operator and calibrating the initial wave head of the fault traveling wave data according to the signal energy spectrum of the highest frequency traveling wave modal component signal include the following steps: converting all of the traveling wave modal component signals into discrete component signals; The signal energy spectrum of each discrete component signal is calculated using the NTEO energy operator. The calculation formula of the signal energy spectrum is as follows: Where: Ψ[h i (x)] represents the i-th discrete component signal h i (x) signal energy spectrum, T represents the resolution parameter, M represents the signal number of the discrete component signal; The signal energy spectrum corresponding to the traveling wave modal component signal in the highest frequency domain is selected as the target signal energy spectrum, and the time point with the highest energy value in the target signal energy spectrum is calibrated as the initial wave head of the fault traveling wave data.

7. The fault monitoring and management method for a 35kV wind farm overhead collector line according to claim 2, characterized in that: The fault traveling wave component includes a fault traveling wave line mode component and a fault traveling wave zero mode component, the fault traveling wave line mode component and the fault traveling wave zero mode component share the fault traveling wave initial wave head, the fault coefficients of all the collector branch lines are calculated based on the fault traveling wave initial wave head and according to the collector line topology structure, and the line fault point in the collector line topology structure is located in combination with all the fault coefficients, including the following steps: The fault data detection terminal node closest to the wind farm power transmission system node in the collector line topology is used as the main line starting node, and the fault data detection terminal node farthest from the wind farm power transmission system node is used as the main line terminating node; The direction from the starting node of the main line to the ending node of the main line is defined as the positive direction of the traveling wave; Based on the initial wave head of the fault traveling wave and the first propagation velocity of the fault traveling wave line mode component, respectively calculating a first propagation time for the fault traveling wave line mode component to propagate to the starting node of the main line and a second propagation time for the fault traveling wave line mode component to propagate to the ending node of the main line; Based on the initial wave head of the fault traveling wave and the second propagation velocity of the zero-mode component of the fault traveling wave, respectively calculating a third propagation time for the zero-mode component of the fault traveling wave to propagate to the starting node of the main line and a fourth propagation time for the zero-mode component of the fault traveling wave to propagate to the ending node of the main line; Calculate, by combining the first propagation time, the second propagation time, the third propagation time, the fourth propagation time, and the first propagation speed, a first fault distance between an imaginary fault point and a starting node of the main line, and a second fault distance between the imaginary fault point and a terminating node of the main line; Combining the first fault distance, the second fault distance, and the node edge attributes in the collector line topology structure, and calculating according to the positive direction of the traveling wave to obtain the fault coefficients of all the collector branch lines; If there is any collector branch line with a fault coefficient of 0, it is determined that the line fault point generating the fault traveling wave data is located on the fault collector branch line with a fault coefficient of 0, and the midpoint of the fault collector branch line is used as the line fault point in the collector line topology structure; If all the fault coefficients are not 0, it is determined that the line fault point is located in the main collector line. The initial wave head of the fault traveling wave, the propagation speed of the fault traveling wave component and the node edge attribute are combined for calculation, and the double-terminal traveling wave positioning method is used to determine the line fault point in the collector line topology structure.

8. The fault monitoring and management method for a 35kV wind farm overhead collector line according to claim 7, characterized in that: The calculation formula of the first fault distance is as follows: Where: represents the first fault distance between the hypothetical fault point G and the main line starting node Q1, t1 represents the first propagation time, t2 represents the second propagation time, t3 represents the third propagation time, t4 represents the fourth propagation time, and v1 represents the first propagation speed; The calculation formula of the second fault distance is as follows: Where: Indicates the hypothetical fault point G and the main line termination node Q n The second fault distance between 9. The fault monitoring and management method for a 35kV wind farm overhead collector line according to claim 1, characterized in that: Extracting energy spectrum features of all signal energy spectra, and identifying the fault type of the line fault point through a deep learning model based on the energy spectrum features includes the following steps: For each of the traveling wave modal component signal's signal energy spectrum, energy spectrum segments of different time lengths are cut out from the signal energy spectrum through time windows of different time scales, and energy time domain features are extracted from each of the energy spectrum segments, wherein the energy time domain features include energy peak value, energy decay rate, energy rise time, energy duration, and energy phase difference feature; Extracting energy frequency domain features of the signal energy spectrum in the frequency domain using a statistical method, wherein the energy frequency domain features include energy proportion, energy mean, energy variance and energy kurtosis; Using a correlation analysis method, a subset of energy features capable of distinguishing fault types is selected from the energy time domain features and the energy frequency domain features; The energy feature subset is input into a pre-trained energy feature recognition model, and the fault type of the line fault point is output through the energy feature recognition model. The energy feature recognition model is a convolutional neural network model including a multi-head attention mechanism.

10. A fault monitoring and management system for a 35kV wind farm overhead collector line, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the fault monitoring and management method for the 35kV wind farm overhead collector line according to any one of claims 1 to 9 is implemented.

Images