Intelligent distribution network fault detection and diagnosis method and system

Abstract

The invention discloses an intelligent distribution network fault detection and diagnosis method and system, which can more accurately calculate the position of a fault point by capturing the wave head of a fault traveling wave by adopting high-frequency sampling and extracting the arrival time of the traveling wave by utilizing a double-end time difference measurement positioning technology and combining wavelet transform. According to the method, the fault position can be quickly and accurately determined, the power failure time and the maintenance cost are reduced, the nonlinear characteristics of the arc fault are identified by calculating the high-frequency band energy entropy value, the arc fault which is difficult to find by a traditional method can be effectively detected, and the safety and the reliability of power grid operation are improved.

Description

Technical Field

[0001] The present invention relates to the field of distribution network detection, and particularly to an intelligent distribution network fault detection and diagnosis method and system. Background Art

[0002] In modern power systems, ensuring the safe and stable operation of the distribution network is of crucial importance. However, traditional fault location techniques often struggle to accurately determine the specific location of a fault when faced with complex network structures. Especially in the case of arc faults with non-linear characteristics, traditional methods prove inadequate. Due to the transient and non-linear nature of arc faults, accurately identifying and locating them using conventional means poses a significant challenge. This not only affects the speed of fault handling but also may lead to increased blindness in maintenance work, thereby prolonging the power outage time and affecting the user's power consumption experience.

[0003] With the continuous expansion of the power grid scale and technological advancements, the grid structure and operating environment have become increasingly complex and variable. New devices are constantly being connected, and the proportion of distributed energy sources is increasing year by year. These changes have placed higher requirements on existing fault detection and diagnosis methods. Regrettably, current methods show significant lag in adapting to these new conditions or fault patterns. They may not be able to quickly respond to changes in the grid structure or the emergence of new fault types, thus affecting the safety and reliability of the entire system. For example, in some areas with a high proportion of renewable energy access, traditional methods may not be able to effectively identify new fault patterns caused by the intermittency and volatility of new energy.

[0004] In summary, there is a need for an intelligent distribution network fault detection and diagnosis method and system to address the deficiencies in the existing technology. Summary of the Invention

[0005] Aiming at the deficiencies of the existing technology, the present invention provides an intelligent distribution network fault detection and diagnosis method and system to address the above problems.

[0006] To achieve the above object, the present invention provides the following technical solution: An intelligent distribution network fault detection and diagnosis method, comprising the following steps: Step S1: Configure multi-source sensing devices at key nodes of the distribution network, and construct a communication network architecture to obtain compliance curves and alarm information from the distribution automation system and the advanced metering infrastructure; Step S2: Capture the wavefront of the fault traveling wave by high-frequency sampling, locate by double-ended time difference measurement, extract the arrival time of the traveling wave by wavelet transform, calculate the energy entropy value in the high-frequency band, and identify the non-linear characteristics of the arc fault; Step S3: Detect the over-limit of current / voltage amplitude, calculate the impedance value and compare it with the preset threshold to determine the fault section. Integrate PMU and SCADA for multi-source information fusion, construct a relationship diagram of distribution network equipment, and output the comprehensive diagnosis result; Step S4: Determine the fault section through the fault indication information of adjacent nodes, determine the fault section through the consistency algorithm, combine with the graph model established for the distribution network topology, mark the positions of sectional switches, and output the optimal fault point coordinates based on the optimized objective function algorithm. Summarize the results to generate a positioning topology map; Step S5: Inject faults into the RTDS real-time digital simulation model to verify the accuracy of the algorithm and output a fault report.

[0007] Optionally, step S2 is implemented in the following manner: Step A1: High-frequency sampling and data acquisition, perform high-frequency sampling on the current or voltage signals at both ends of the power line to obtain the acquisition data set; Step A2: Double-end time difference measurement for positioning, record the time when the fault traveling wave arrives at both ends of the line through double-end measurement, and calculate the fault point position using the traveling wave propagation speed and the line length; Step A3: Signal transformation to extract the arrival time of the traveling wave, transform the acquired high-frequency signal, extract the arrival time of the traveling wave head, and obtain the arrival time of the traveling wave head by analyzing the position of the maximum value of the wavelet coefficient; Step A4: Calculate the energy entropy value of the high-frequency band, calculate the energy entropy value of the high-frequency components after wavelet decomposition, and identify the non-linear characteristics of the arc fault; Step A5: According to the change trend of the energy entropy value of the high-frequency band, combined with the time-domain characteristics extracted by wavelet transform, determine whether there is an arc fault. If the energy entropy value is higher than the threshold in the normal operating state, it is determined as an arc fault.

[0008] Optionally, the transformation formula in step A3 is:

[0009] where \(W(a,b)\) is the wavelet coefficient, \(k\) is the conversion coefficient, \(f(t)\) is the original signal, \(\psi\) is the wavelet basis function, \(a\) is the scaling parameter, \(b\) is the translation parameter, and \(\delta\) is the error term.

[0010] Optionally, the energy entropy in step A4 is calculated in the following manner: , , where \(p\) i is the energy proportion of the \(i\)-th frequency band, \(E\) i is the energy of the \(i\)-th frequency band, and \(N\) is the total number of frequency bands.

[0011] Optionally, step S3 is implemented as follows: Step B1: Data acquisition, collect voltage and current phasor data in real time, and obtain discrete event information such as switch status and protection action signals; Step B2: Magnitude over-limit detection, for each line, set the safe upper and lower limits of current and voltage, and compare whether the real-time detected current and voltage values exceed the preset upper and lower limits; Step B3: Impedance calculation and comparison, use Ohm's law to calculate the line impedance, and compare the calculated impedance value with the impedance threshold within the predefined normal operating range to determine whether it is abnormal; Step B4: Based on the impedance calculation result and the phase angle difference information provided by the PMU, locate the fault position by calculation, and combine the switch status information of the SCADA system to assist in determining the specific fault section; Step B5: Integrate the continuous time series data provided by the PMU and the discontinuous event information provided by the SCADA, process the fused data, and accurately identify the fault; Step B6: Construct a distribution network equipment relationship diagram, establish a connection relationship diagram between distribution network equipment according to the network topology structure, and mark the fault occurrence position and its influence range in the relationship diagram; Step B7: Comprehensive analysis results, generate a report including fault type, location, severity, and recommended operation measures.

[0012] Optionally, step S4 is implemented as follows: Step C1: Construct a graph model according to the physical connection relationship of the distribution system; Step C2: Fault indication information analysis, collect fault indication information from each node, and for each node, record whether it detects a fault signal; Step C3: Based on the consensus algorithm of message passing, let adjacent nodes exchange fault indication information, gradually converge to a consistent fault area judgment, and update its own state by weighted average or logical operation of the fault states of neighbor nodes; Step C4: Mark all sectional switch positions on the graph model, define an optimization objective function, use numerical optimization methods to solve the objective function, obtain the optimal fault point coordinates, map the obtained fault point coordinates back to the distribution network topology graph model, mark the specific fault position, and use a graphical tool to draw the final positioning topology graph to show the fault point and its influence range; Step C5: Summarize all analysis results, including but not limited to the exact position of the fault point, the affected area, and recommended repair measures, and generate a detailed report document.

[0013] Optionally, step S5 is implemented as follows: Step D1: Establish a simulation model, build a detailed topological model of the distribution network in RTDS, and make the model parameters consistent with the actual system; Step D2: Inject a fault and collect data. Use the fault injection function of RTDS to inject a preset fault signal into the model at a specified time point, export the time series data of key electrical quantities from RTDS, and record the fault indication information of each node and branch; Step D3: Apply an algorithm for fault location. Use the collected data as input and pass it to the fault location algorithm. Update the node status based on the consistency algorithm and gradually converge to a consistent fault area judgment; Step D4: Define an objective function to minimize the weighted sum of the distances between the fault point and the fault indication nodes, use an optimization algorithm to solve the optimal fault point coordinates, verify the correctness of the algorithm, and output a fault report.

[0014] An intelligent distribution network fault detection and diagnosis system, which adopts the intelligent distribution network fault detection and diagnosis method, is characterized in that it includes a multi-source sensing device configuration module, a high-frequency sampling and fault traveling wave analysis module, an amplitude limit detection and impedance comparison module, a multi-source information fusion and comprehensive diagnosis module, a fault section determination and optimization location module, and an RTDS real-time digital simulation verification module; The multi-source sensing device configuration module is responsible for installing and configuring multi-source sensing devices at key nodes of the distribution network and constructing a communication network architecture; The high-frequency sampling and fault traveling wave analysis module is used for high-frequency sampling and fault traveling wave analysis of the distribution network; The amplitude limit detection and impedance comparison module is used for detecting the amplitude limit of current and voltage and comparing the impedance; The multi-source information fusion and comprehensive diagnosis module is used for fusing multi-source information by combining the data provided by PMU and SCADA, constructing a relationship diagram of distribution network equipment, and outputting a comprehensive diagnosis result; The fault section determination and optimization location module is used for determining the fault section based on the consistency algorithm and optimizing the location; The RTDS real-time digital simulation verification module is used for verifying by establishing a detailed distribution network simulation model, injecting a fault in RTDS, and applying the aforementioned algorithm, and outputting a fault report.

[0015] Optionally, the high-frequency sampling and fault traveling wave analysis module includes a high-frequency sampling unit, a double-end time difference measurement and location unit, and a wavelet transform unit; The high-frequency sampling unit is used for high-frequency sampling of the current or voltage signals at both ends of the power line; The double-end time difference measurement and location unit is used for calculating the fault point location by recording the time when the fault traveling wave arrives at both ends of the line through double-end measurement; The wavelet transform unit is used to extract the arrival time of traveling waves, calculate the energy entropy value in the high-frequency band, and identify the non-linear characteristics of arc faults.

[0016] Optionally, the fault section determination and optimization positioning module includes a fault indication information analysis unit, a consensus algorithm unit, and a numerical optimization solution unit; The fault indication information analysis unit is used to collect fault indication information from each node; The consensus algorithm unit is used to gradually converge to a consistent fault area judgment by exchanging fault indication information between adjacent nodes; The numerical optimization solution unit is used to define an optimization objective function, solve for the optimal fault point coordinates using numerical optimization methods, and generate a positioning topology map.

[0017] Advantages of the present invention: 1. In the present invention, by using high-frequency sampling to capture the wavefront of fault traveling waves, and using the double-ended time difference positioning technology, combined with wavelet transform to extract the arrival time of traveling waves, the position of the fault point can be calculated more accurately. This helps to quickly and accurately determine the fault location, reduce power outage time and maintenance costs. By calculating the energy entropy value in the high-frequency band to identify the non-linear characteristics of arc faults, this method can effectively detect arc faults that are difficult to discover by traditional methods, improving the safety and reliability of power grid operation; 2. In the present invention, multi-source information provided by PMU and SCADA is fused to construct a relationship map of distribution network equipment. This comprehensive analysis method can perform fault diagnosis more comprehensively and accurately, provide a more detailed fault information report. Based on the consensus algorithm and numerical optimization solution method, the fault section can be determined more efficiently and accurately in a complex distribution network, and the optimal fault point coordinates can be output, generating a positioning topology map, so as to guide on-site staff to take measures quickly; 3. In the present invention, by building a detailed distribution network model in RTDS and injecting preset fault signals, the proposed fault detection and diagnosis algorithm can be comprehensively verified to ensure its reliability and effectiveness in practical applications. The modular design is adopted, including a multi-source sensing device configuration module, a high-frequency sampling and fault traveling wave analysis module, etc., making the entire system have high flexibility and scalability, and can better adapt to changes in power grid structure and new operating conditions. Brief Description of the Drawings

[0018] Figure 1 It is a schematic diagram of a method flow of the present invention.

[0019] Figure 2 It is a flowchart of step S2 of the present invention.

[0020] Figure 3It is a flowchart of step S3 of the present invention.

[0021] Figure 4 It is a flowchart of step S4 of the present invention.

[0022] Figure 5 It is a flowchart of step S5 of the present invention.

[0023] Figure 6 It is a schematic diagram of the system structure of the present invention. Detailed implementation manners

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the invention or the prior art, the following will briefly introduce the accompanying drawings required for use in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the invention. For those of ordinary skill in the art, other accompanying drawings can be obtained based on these drawings without creative efforts.

[0025] As Figures 1 to 5 shown, an intelligent distribution network fault detection and diagnosis method includes the following steps: Step S1: Configure multi-source sensing devices at key nodes of the distribution network, and construct a communication network architecture to obtain compliance curves and alarm information from the distribution automation system and the advanced metering infrastructure; Step S2: Capture the fault traveling wave head by high-frequency sampling, locate it by double-ended time difference measurement, extract the arrival time of the traveling wave through wavelet transform, and calculate the energy entropy value in the high-frequency band to identify the non-linear characteristics of the arc fault, specifically through the following methods: 1. High-frequency sampling and data acquisition Use high-speed data acquisition equipment to perform high-frequency sampling on the current or voltage signals at both ends of the power line (the sampling frequency is usually at the MHz level).

[0026] The acquired data includes the fault traveling wave head signal.

[0027] 2. Double-ended time difference positioning Through double-ended measurement technology, record the time t1 and t2 when the fault traveling wave arrives at both ends of the line.

[0028] Use the traveling wave propagation speed V and the line length L to calculate the fault point position through the following formula: , In the formula, t1 is the time when the traveling wave arrives at section A, t2 is the time when the traveling wave arrives at end B, and v is the traveling wave propagation speed; 3. Wavelet transform to extract the arrival time of the traveling wave Perform wavelet transform on the acquired high-frequency signal to extract the arrival time of the traveling wave head.

[0029] Use the continuous wavelet transform (CWT) formula: , where W(a, b) is the wavelet coefficient, k is the conversion coefficient, f(t) is the original signal, is the wavelet basis function, a is the scaling parameter, b is the translation parameter, and δ is the error term; 4. Calculate the energy entropy value of the high-frequency band Calculate the energy entropy value of the high-frequency components after wavelet decomposition to identify the non-linear characteristics of the arc fault.

[0030] The energy entropy is calculated as follows: , , where pi is the energy proportion of the i-th frequency band, Ei is the energy of the i-th frequency band, and N is the total number of frequency bands.

[0031] The larger the energy entropy value of the high-frequency band, the more obvious the non-linear characteristics of the signal.

[0032] 5. Fault feature identification Based on the change trend of the energy entropy value of the high-frequency band and combined with the time-domain features extracted by wavelet transform, judge whether there is an arc fault.

[0033] If the energy entropy value is significantly higher than the threshold in the normal operating state, it is determined as an arc fault.

[0034] Step S3: Detect the over-limit of current / voltage amplitude, calculate the impedance value and compare it with the preset threshold, determine the fault section, fuse the PMU and SCADA for multi-source information fusion, construct a relationship diagram of distribution network equipment, and output the comprehensive diagnosis result, specifically in the following way: 1. Data acquisition PMU: Collect real-time high-precision voltage and current phasor data, including amplitude and phase.

[0035] SCADA: Obtain discrete event information such as switch status and protection action signals.

[0036] 2. Amplitude over-limit detection For each line, set the safe upper and lower limits of current and voltage (Imin, Imax and Vmin, Vmax).

[0037] Compare whether the real-time monitored current and voltage values exceed these preset upper and lower limits: If I < Imin or I > Imax or V < Vmin or V > Vmax, an alarm is triggered.

[0038] If I < Imin or I > Imax or V < Vmin or V > Vmax, an alarm is triggered.

[0039] 3. Impedance Calculation and Comparison Use Ohm's law to calculate the line impedance Z: Z = V / I, Compare the calculated impedance value with the impedance threshold within the predefined normal operating range to determine if it is abnormal.

[0040] 4. Fault Section Determination Based on the impedance calculation results and the phase angle difference information provided by the PMU, use power system analysis methods such as power flow analysis or short - circuit calculation to locate the possible fault positions.

[0041] Combine the switch status information of the SCADA system to assist in determining the specific fault section.

[0042] 5. Multi - source Information Fusion Integrate the continuous time - series data (such as voltage and current phasors) provided by the PMU with the discontinuous event information (such as switch tripping) provided by the SCADA.

[0043] Use data mining techniques or machine learning algorithms to process the fused data to improve the accuracy of fault identification.

[0044] 6. Constructing the Distribution Network Equipment Relationship Diagram Establish the connection relationship diagram between distribution network equipment according to the network topology structure.

[0045] Mark the location of the fault and its influence range in the relationship diagram.

[0046] 7. Outputting the Comprehensive Diagnosis Result Based on all the above analysis results, generate a report including the fault type, location, severity, and recommended operation measures. Step S4: Through the fault indication information of adjacent nodes, determine the fault section through the consistency algorithm, combine with the graph model of the distribution network topology, mark the position of the sectionalizing switch, based on the optimization objective function algorithm, output the optimal fault point coordinates, summarize the results to generate the positioning topology diagram, and implement it in the following ways: 1. Data Preparation and Network Modeling Data collection: Obtain electrical parameters including voltage and current and switch status information from the SCADA system; obtain high - precision phasor measurement data from the PMU.

[0047] Network modeling: Construct a graph model G=(V, E) according to the physical connection relationship of the distribution system, where Let \(V\) represent the set of nodes (such as substations, branch points, etc.), and \(E\) represent the set of edges (such as transmission lines). Each node and edge should contain its attribute information, such as capacity, impedance, etc.

[0048] 2. Fault Indication Information Analysis Collect fault indication information from each node, which may come from devices such as overcurrent protection devices and fault indicators.

[0049] For each node \(v_i\in V\), record whether it detects a fault signal. Assume that if node \(v_i\) detects a fault, then mark it as \(F(v_i)=1\); otherwise, \(F(v_i)=0\).

[0050] 3. Use the Consensus Algorithm to Determine the Fault Section Apply a distributed consensus algorithm on the graph model \(G\), such as a consensus algorithm based on message passing, to let adjacent nodes exchange fault indication information and gradually converge to a consistent fault area judgment.

[0051] The calculation formula can be determined according to the specific consensus algorithm, but usually involves weighted averaging or logical operations on the fault states of neighbor nodes to update its own state.

[0052] 4. Mark the Positions of Sectionalizing Switches Clearly mark the positions of all sectionalizing switches in the graph model. This step is crucial for subsequent fault isolation.

[0053] The sectionalizing switches can be regarded as special edges and marked with different symbols or colors in the graph model.

[0054] 5. Calculate the Optimal Fault Point Coordinates Based on the Optimization Objective Function Define an optimization objective function that takes into account factors such as minimizing the sum of the distances from the fault point to each fault indication node.

[0055] The objective function can be expressed as: , where \(w_i\) is the weight coefficient, which can be set according to the importance of different nodes; is the distance from the fault point to node \(v_i\).

[0056] 6. Output the Optimal Fault Point Coordinates and Generate the Location Topology Map Use numerical optimization methods to solve the above objective function to obtain the optimal fault point coordinates.

[0057] Map the obtained fault point coordinates back to the distribution network topology graph model and mark the specific fault location.

[0058] Use a graphical tool to draw the final location topology diagram, clearly showing the fault point and its affected range.

[0059] 7. Result Summarization and Report Generation Summarize all the analysis results, including the exact location of the fault point, the affected area, the recommended repair measures, etc.

[0060] Step S5: Inject a fault through the RTDS real-time digital simulation model to verify the algorithm accuracy and output a fault report. The specific implementation is as follows: 1. Preparation Phase Build a simulation model: Build a detailed topology model of the distribution network in RTDS, including power sources, lines, loads, switching equipment, etc.

[0061] Ensure that the model parameters (such as line impedance, transformer turns ratio, load characteristics, etc.) are consistent with the actual system.

[0062] 2. Inject Fault and Collect Data Inject Fault: Use the fault injection function of RTDS to inject a preset fault signal into the model at a specified time point. For example: , where is the normal operating voltage, is the fault voltage, is the fault start time, is the fault end time.

[0063] Collect Data: Export the time series data of key electrical quantities (such as voltage, current, power, etc.) from RTDS.

[0064] Record the fault indication information (such as overcurrent signal, zero-sequence current, etc.) of each node and branch.

[0065] 3. Apply the Algorithm for Fault Location Input Data Processing: Use the collected data as input and pass it to the fault location algorithm.

[0066] Perform necessary preprocessing on the input data, such as filtering, normalization, etc.

[0067] Apply the Consistency Algorithm: Update the node states based on the consistency algorithm and gradually converge to a consistent fault area judgment.

[0068] The calculation formula is: , where $s_{i,k}$ is the state value of the $i$-th node at the $k$-th iteration, $N(i)$ is the set of neighbors of node $i$, and $\alpha$ is the step size coefficient.

[0069] Optimization objective function solving: Define the objective function to minimize the weighted sum of the distances between the fault points and the fault indication nodes, use an optimization algorithm to solve the optimal fault point coordinates, verify the correctness of the algorithm, and output a fault report.

[0070] The present invention can capture the fault traveling wave head by high-frequency sampling, use the double-ended time difference location technology, and combine wavelet transform to extract the arrival time of the traveling wave, so as to calculate the position of the fault point more accurately. This helps to quickly and accurately determine the fault location, reduce the power outage time and maintenance costs. By calculating the energy entropy value in the high-frequency band to identify the nonlinear characteristics of arc faults, this method can effectively detect arc faults that are difficult to discover by traditional methods, improving the safety and reliability of power grid operation; Integrates multi-source information provided by PMU and SCADA to construct a distribution network equipment relationship diagram. This comprehensive analysis method can perform fault diagnosis more comprehensively and accurately, provide a more detailed fault information report. Based on the consensus algorithm and numerical optimization solution method, it can more efficiently and accurately determine the fault section in a complex distribution network, output the optimal fault point coordinates, and generate a location topology diagram, so as to guide on-site staff to take measures quickly; By building a detailed distribution network model in RTDS and injecting preset fault signals, the proposed fault detection and diagnosis algorithm can be comprehensively verified to ensure its reliability and effectiveness in practical applications. It adopts a modular design, including a multi-source sensing device configuration module, a high-frequency sampling and fault traveling wave analysis module, etc., making the entire system have high flexibility and scalability, and can better adapt to the changes in the power grid structure and new operating conditions.

[0071] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, or improvements made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent distribution network fault detection and diagnosis method, characterized in that, It includes the following steps: Step S1: Configure multi-source sensing devices at key nodes of the distribution network, build a communication network architecture, and obtain compliance curves and alarm information from the distribution automation system and the advanced metering infrastructure; Step S2: Use high-frequency sampling to capture the wavefront of the fault traveling wave, locate by double-ended time difference measurement, extract the arrival time of the traveling wave through wavelet transform, calculate the energy entropy value in the high-frequency band, and identify the non-linear characteristics of the arc fault; Step S3: Detect current / voltage amplitude over-limit, calculate the impedance value and compare it with a preset threshold to determine the fault section, fuse PMU and SCADA for multi-source information fusion, build a relationship diagram of distribution network devices, and output the comprehensive diagnosis result; Step S4: Determine the fault section through the fault indication information of adjacent nodes, combine with the graph model established for the distribution network topology, mark the position of the sectionalizing switch, and based on the optimization objective function algorithm, output the optimal fault point coordinates, and summarize the results to generate a positioning topology map; Step S5: Inject faults into the RTDS real-time digital simulation model to verify the accuracy of the algorithm and output a fault report.

2. The intelligent distribution network fault detection and diagnosis method according to claim 1, characterized in that, The implementation of the said step S2 is as follows: Step A1: High-frequency sampling and data acquisition, perform high-frequency sampling on the current or voltage signals at both ends of the power line to obtain the acquisition data set; Step A2: Double-ended time difference positioning, through double-ended measurement, record the time when the fault traveling wave arrives at both ends of the line, and calculate the fault point position using the traveling wave propagation speed and the line length; Step A3: Signal transformation to extract the arrival time of the traveling wave, transform the acquired high-frequency signal, extract the arrival time of the traveling wavefront, and obtain the arrival time of the traveling wavefront by analyzing the position of the maximum value of the wavelet coefficient; Step A4: Calculate the energy entropy value in the high-frequency band, calculate the energy entropy value for the high-frequency components after wavelet decomposition, and identify the non-linear characteristics of the arc fault; Step A5: According to the change trend of the energy entropy value in the high-frequency band, combined with the time-domain characteristics extracted by wavelet transform, judge whether there is an arc fault. If the energy entropy value is higher than the threshold in the normal operating state, it is determined as an arc fault.

3. The intelligent distribution network fault detection and diagnosis method according to claim 2, characterized in that, The transformation formula in the said step A3 is: ， Wherein, W(a, b) is the wavelet coefficient, k is the conversion coefficient, f(t) is the original signal, is the wavelet basis function, a is the scaling parameter, b is the translation parameter, and δ is the error term.

4. The intelligent distribution network fault detection and diagnosis method according to claim 2, wherein, The energy entropy in the said step A4 is calculated in the following way: ， ， where p i is the energy proportion of the i-th frequency band, E i is the energy of the i-th frequency band, and N is the total number of frequency bands.

5. The intelligent distribution network fault detection and diagnosis method according to claim 1, wherein The implementation of the said step S3 is as follows: Step B1: Data acquisition, collect voltage and current phasor data in real time, and obtain discrete event information such as switch status and protection action signals; Step B2: Amplitude over-limit detection, for each line, set the safe upper and lower limits of current and voltage, and compare whether the real-time detected current and voltage values exceed the preset upper and lower limits; Step B3: Impedance calculation and comparison, calculate the line impedance using Ohm's law, and compare the calculated impedance value with the impedance threshold within the predefined normal operating range to judge whether it is abnormal; Step B4: Based on the impedance calculation result and the phase angle difference information provided by the PMU, calculate and locate the fault position, and combine the switch status information of the SCADA system to assist in determining the specific fault section; Step B5: Integrate the continuous time series data provided by the PMU and the discontinuous event information provided by the SCADA, process the fused data, and accurately identify the fault; Step B6: Construct a distribution network equipment relationship diagram, establish a connection relationship diagram between distribution network equipment according to the network topology structure, and mark the location where the fault occurs and its influence range in the relationship diagram; Step B7: Synthesize the analysis results and generate a report including the fault type, location, severity, and recommended operation measures.

6. The intelligent distribution network fault detection and diagnosis method according to claim 1, characterized in that, The said Step S4 is implemented in the following manner: Step C1: Construct a graph model according to the physical connection relationship of the distribution system; Step C2: Analyze fault indication information, collect fault indication information from each node, and for each node, record whether it detects a fault signal; Step C3: Based on the message-passing consistency algorithm, let adjacent nodes exchange fault indication information, gradually converge to a consistent fault area judgment, and update their own states through weighted averaging or logical operations on the fault states of neighbor nodes; Step C4: Mark the positions of all sectionalizing switches on the graph model, define an optimization objective function, use numerical optimization methods to solve the objective function, obtain the optimal fault point coordinates, map the obtained fault point coordinates back to the distribution network topology graph model, mark the specific fault location, and use a graphical tool to draw the final positioning topology graph to show the fault point and its influence range; Step C5: Summarize all the analysis results, including but not limited to the exact location of the fault point, the affected area, and the recommended repair measures, and generate a detailed report document.

7. The intelligent distribution network fault detection and diagnosis method according to claim 1, wherein The said Step S5 is implemented in the following manner: Step D1: Establish a simulation model, build a detailed topology model of the distribution network in RTDS, and make the model parameters consistent with the actual system; Step D2: Inject a fault and collect data, use the fault injection function of RTDS to inject a preset fault signal into the model at a specified time point, export the time series data of key electrical quantities from RTDS, and record the fault indication information of each node and branch; Step D3: Apply an algorithm for fault location, use the collected data as input, transfer it to the fault location algorithm, update the node states based on the consistency algorithm, and gradually converge to a consistent fault area judgment; Step D4: Define an objective function to minimize the weighted sum of the distances between the fault point and the fault indication nodes, use an optimization algorithm to solve the optimal fault point coordinates, verify the correctness of the algorithm, and output a fault report.

8. An intelligent distribution network fault detection and diagnosis system, which adopts the intelligent distribution network fault detection and diagnosis method according to any one of claims 1-7, is characterized in that, Including a multi-source sensing device configuration module, a high-frequency sampling and fault traveling wave analysis module, an amplitude limit detection and impedance comparison module, a multi-source information fusion and comprehensive diagnosis module, a fault section determination and optimal positioning module, and an RTDS real-time digital simulation verification module; The multi-source sensing device configuration module is responsible for installing and configuring multi-source sensing devices at key nodes of the distribution network and constructing a communication network architecture; The high-frequency sampling and fault traveling wave analysis module is used for high-frequency sampling and fault traveling wave analysis of the distribution network; The amplitude limit detection and impedance comparison module is used for detecting current and voltage amplitude limits and comparing impedances; The multi-source information fusion and comprehensive diagnosis module is used to combine the data provided by PMU and SCADA, perform multi-source information fusion, construct a distribution network equipment relationship diagram, and output comprehensive diagnosis results; The fault section determination and optimization positioning module is used to determine the fault section based on the consistency algorithm and optimize the positioning; The RTDS real-time digital simulation verification module is used to establish a detailed distribution network simulation model, inject faults in RTDS, verify using the aforementioned algorithm, and output a fault report.

9. The intelligent distribution network fault detection and diagnosis system according to claim 8, characterized in that, The high-frequency sampling and fault traveling wave analysis module includes a high-frequency sampling unit, a double-ended time difference measurement positioning unit, and a wavelet transform unit; The high-frequency sampling unit is used to perform high-frequency sampling on the current or voltage signals at both ends of the power line; The double-ended time difference measurement positioning unit is used to calculate the fault point location by recording the arrival time of the fault traveling wave at both ends of the line through double-ended measurement; The wavelet transform unit is used to extract the arrival time of the traveling wave, calculate the high-frequency band energy entropy value, and identify the nonlinear characteristics of the arc fault.

10. The intelligent distribution network fault detection and diagnosis system according to claim 8, wherein The fault section determination and optimization positioning module includes a fault indication information analysis unit, a consistency algorithm unit, and a numerical optimization solution unit; The fault indication information analysis unit is used to collect fault indication information from each node; The consistency algorithm unit is used to gradually converge to a consistent fault area judgment by exchanging fault indication information between adjacent nodes; The numerical optimization solution unit is used to define an optimization objective function, solve for the optimal fault point coordinates using numerical optimization methods, and generate a positioning topology map.

Images