Method, system, device and medium for quickly locating faults based on three-stage protection and FA coordinated protection

Abstract

The present invention discloses a method, system, device, and medium for rapidly locating faults based on three-stage protection and FA collaborative protection. The method specifically includes: using a preset three-stage action rule for detection and judgment to obtain a real-time protection action signal; real-time acquisition of action timing data of each protection device in the distribution network, constructing an FA fault feature matrix based on the action timing data, electrical quantity information, and topological structure, and using an LSTM neural network to analyze the temporal correlation of the FA fault feature matrix to generate an FA fault probability distribution map; based on the real-time protection action signal and the FA fault probability distribution map, using a fault location algorithm for collaborative analysis to obtain a fault location result; based on the fault location result, using a reinforcement learning algorithm to dynamically plan the load transfer path of the distribution network and generate an optimal isolation and restoration plan. The present invention is applicable to various complex distribution network structures and operating modes and can achieve rapid and accurate fault location and processing under different fault conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution networks, and in particular to a method, system, equipment and medium for quickly locating faults based on three-stage protection and FA collaborative protection. Background Art

[0002] In power systems, rapid and accurate fault location, followed by isolation and restoration of power, is critical to ensuring power supply reliability and power quality. Traditionally, the three-stage protection mechanism, comprising current quick-trip protection, time-limited current quick-trip protection, and overcurrent protection, has served as the foundational protection method in power systems and has achieved a certain degree of fault detection and response. However, with the increasing complexity of distribution networks, particularly those with numerous branches and rings, the three-stage protection mechanism has gradually revealed its limitations in terms of the accuracy and speed of fault location.

[0003] Specifically, the application of traditional three-stage protection in distribution networks faces numerous challenges. First, quick-trip protection often struggles to coordinate perfectly with higher-level protection, leading to overlapping or missing protection ranges, impacting protection accuracy and reliability. Second, while overcurrent protection can cover a wider range of faults, its long operating delay often fails to meet the rapid fault handling requirements of modern power systems.

[0004] At the same time, feeder automation (FA) technology, as an advanced means of monitoring and controlling distribution networks, significantly improves the automation level of distribution networks by monitoring their operating status in real time and incorporating intelligent algorithms to locate, isolate, and restore power to faults. However, standalone FA systems can also have shortcomings in certain situations. For example, when fault signals are complex or multiple fault points exist, the FA system may be unable to accurately and promptly determine the fault location. Furthermore, the performance of the FA system is heavily dependent on the reliability and speed of the communication network. Failures or delays in the communication network can severely impact the performance of the FA system.

[0005] To overcome the limitations of traditional three-stage protection and standalone FA systems, several improvements have been proposed. For example, attempts have been made to improve protection accuracy and reliability by introducing current-based criteria and voltage-assisted blocking mechanisms. However, these solutions do not fundamentally address the technical challenges of adaptively adjusting protection settings and linking FA logic. In complex topologies, such as those involving four-segment, three-connection overhead lines, traditional protection methods are prone to over-tripping and other issues, further impacting the stable operation of the distribution network. Summary of the Invention

[0006] The purpose of the present invention is to provide a method, system, device and medium for quickly locating faults based on three-stage protection and FA collaborative protection. By real-time acquisition and preprocessing of electrical quantity information and coordinating three-stage protection and FA technology, faults can be quickly located, thereby improving the fault handling efficiency and reliability of the distribution network. The method is applicable to various complex distribution network structures and operating modes, and can achieve rapid and accurate fault location and processing under different fault conditions, thereby solving at least one of the above-mentioned problems of the prior art.

[0007] In a first aspect, the present invention provides a method for quickly locating faults based on three-stage protection and FA collaborative protection, the method specifically comprising:

[0008] Real-time collection of electrical quantity information at each monitoring point in the distribution network, and obtaining the topological structure and real-time operation mode of the distribution network by preprocessing and feature extraction of the electrical quantity information;

[0009] The protection range sensitivity is dynamically calculated based on the topology and real-time operation mode. Based on the protection range sensitivity and electrical quantity information, a preset three-stage action rule is used for detection and judgment to obtain a real-time protection action signal.

[0010] The system collects the action time series data of each protection device in the distribution network in real time, constructs the FA fault feature matrix based on the action time series data, electrical quantity information and topological structure, and uses the LSTM neural network to analyze the time series correlation of the FA fault feature matrix to generate the FA fault probability distribution map.

[0011] Based on the real-time protection action signal and FA fault probability distribution map, the fault location algorithm is used for collaborative analysis to obtain the fault location result;

[0012] Based on the fault location results, a reinforcement learning algorithm is used to dynamically plan the load transfer path of the distribution network and generate the optimal isolation and restoration plan.

[0013] In a second aspect, the present invention provides a system for quickly locating faults based on three-stage protection and FA collaborative protection, the system specifically comprising:

[0014] The data acquisition and preprocessing module is used to collect electrical quantity information of each monitoring point in the distribution network in real time, and obtain the topological structure and real-time operation mode of the distribution network by preprocessing and extracting features of the electrical quantity information;

[0015] The three-stage protection module is used to dynamically calculate the protection range sensitivity based on the topology structure and real-time operation mode. Based on the protection range sensitivity and electrical quantity information, it uses the preset three-stage action rules to perform detection and judgment to obtain real-time protection action signals;

[0016] The FA fault analysis module is used to collect the action time series data of each protection device in the distribution network in real time, build the FA fault feature matrix based on the action time series data, electrical quantity information and topological structure, and use the LSTM neural network to analyze the time series correlation of the FA fault feature matrix to generate the FA fault probability distribution map;

[0017] The collaborative control module is used to perform collaborative analysis based on the real-time protection action signal and the FA fault probability distribution map using the fault location algorithm to obtain the fault location result;

[0018] The fault self-healing module is used to dynamically plan the load transfer path of the distribution network based on the fault location results using a reinforcement learning algorithm to generate the optimal isolation and restoration plan.

[0019] In a third aspect, the present invention provides a computer device comprising: a memory and a processor and a computer program stored in the memory. When the computer program is executed on the processor, it implements a method for quickly locating faults based on three-stage protection and FA collaborative protection as described in any one of the above methods.

[0020] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for quickly locating faults based on three-stage protection and FA collaborative protection as described in any one of the above methods is implemented.

[0021] Compared with the prior art, the present invention has at least one of the following technical effects:

[0022] 1. The present invention quickly locates faults by real-time acquisition and preprocessing of electrical quantity information and coordinating three-stage protection with FA technology, thereby improving the fault handling efficiency and reliability of the distribution network. It is applicable to various complex distribution network structures and operating modes and can achieve rapid and accurate fault location and handling under different fault conditions.

[0023] 2. Based on the accurate fault location results, the present invention can quickly isolate the fault area, and perform load transfer and power supply restoration operations, thereby reducing the power outage time and scope and improving power supply reliability.

[0024] 3. The present invention can more accurately locate the fault position and shorten the fault processing time by collecting electrical quantity information in real time and dynamically calculating the sensitivity of the protection range, combining the three-stage action rules and the time series correlation analysis of the FA fault feature matrix.

[0025] 4. The present invention can dynamically adjust the protection range and set value according to the topological structure and real-time operation mode of the distribution network, effectively solving the problem that traditional protection set values ​​are difficult to adjust adaptively.

[0026] 5. The present invention dynamically plans the load transfer path through the reinforcement learning algorithm and generates the optimal isolation and restoration plan, which can minimize the impact of faults on the operation of the distribution network and improve power supply reliability and power quality.

[0027] 6. The present invention uses wavelet threshold denoising and Lagrange interpolation method to preprocess electrical quantity information, dynamically reconstruct the topology model, and accurately obtain the real-time operation mode of the distribution network.

[0028] 7. The present invention dynamically calculates the sensitivity of the protection range according to the topological structure and real-time operation mode, generates an adaptive three-stage protection criterion set, and improves the accuracy and flexibility of protection.

[0029] 8. The three-stage protection criterion set of the present invention specifies the current threshold, action time and related parameters to ensure the rapidity and reliability of the protection action.

[0030] 9. The present invention uses a two-layer bidirectional LSTM network to analyze the FA fault feature matrix and generate a fault probability distribution map to improve the timing correlation and accuracy of fault location. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0032] Figure 1 This is a flow chart of a method for quickly locating a fault based on three-stage protection and FA collaborative protection provided by an embodiment of the present invention;

[0033] Figure 2 This is a structural diagram of a system for quickly locating faults based on three-stage protection and FA collaborative protection provided by an embodiment of the present invention;

[0034] Figure 3 It is a structural diagram of a computer device provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0035] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0036] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0037] It will also be understood that the term "and / or" used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0038] As used in this specification and the appended claims, the term "if" can be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]," depending on the context.

[0039] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0040] References to "one embodiment" or "some embodiments" in this specification mean that a particular feature, structure, or characteristic described in conjunction with that embodiment is included in one or more embodiments of the present application. Thus, phrases such as "in one embodiment," "in some embodiments," "in other embodiments," and "in other embodiments" appearing in various places in this specification do not necessarily refer to the same embodiment, but rather mean "one or more but not all embodiments," unless otherwise specifically emphasized. The terms "including," "comprising," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0041] In the embodiments of the present application, the execution subject of the process includes a terminal device, which includes but is not limited to: a server, a computer, a smart phone, a tablet computer, and other devices capable of executing the method disclosed in the present application. Figure 1 A flowchart of a method for quickly locating a fault based on three-stage protection and FA coordinated protection disclosed in the first embodiment of the present invention is shown, and is described in detail as follows:

[0042] S101 , collecting electrical quantity information of each monitoring point in the distribution network in real time, and obtaining the topology structure and real-time operation mode of the distribution network by preprocessing and extracting features of the electrical quantity information.

[0043] In this embodiment, current transformers and voltage transformers are installed at key nodes in the distribution network, such as substation outlets and branch line nodes. The collected current and voltage signals are converted into digital signals and transmitted to the data acquisition unit. The data acquisition unit uses a digital filtering algorithm, such as an FIR filter, to filter the collected signals, remove high-frequency noise and interference signals, and verify and filter the electrical quantity data to remove outliers and erroneous data. The data is standardized to meet the requirements of subsequent analysis and processing. For example, a hash table index method can be used to improve the prefix matching speed. By selecting a hash function suitable for power system measurement data, the efficiency of matching prefix strings in the dictionary can be improved, thereby improving the algorithm execution speed.

[0044] Deep learning models (such as convolutional neural networks and recurrent neural networks) are used to extract features from preprocessed electrical quantity data. By training the model, it can automatically capture key features in electrical quantity data, such as voltage fluctuations, current changes, and power factor. For example, deep learning models suitable for processing power grid data can be designed and built. Data processing methods including convolutional neural networks, recurrent neural networks, long short-term memory, autoencoders, and generative adversarial networks can be used to process image data and time series features of power grid equipment, thereby capturing the complex characteristics and patterns of power grid data.

[0045] Based on the extracted electrical quantity characteristics, combined with the physical connections and operating patterns of the distribution network, topological structure analysis is performed. By constructing a formalized foundational model for topological analysis, the topological analysis process is standardized and verified to ensure the accuracy of the analysis results. For example, power system equipment can be selected as the smallest descriptive unit, and the basic electrical topology formed by physical connections through device state information can be extracted and expressed.

[0046] Combine topology analysis results with real-time electrical quantity information to analyze the real-time operation of the distribution network. Evaluate key indicators such as the distribution network's operating status, load conditions, and power factor to promptly identify potential operational risks and issues.

[0047] In this embodiment, real-time electrical quantity information is collected and preprocessed to ensure data accuracy and reliability. Deep learning models are used for feature extraction, improving the efficiency and accuracy of data analysis. Topological structure analysis automatically builds and updates the distribution network topology model, providing accurate topological information support for distribution network planning, design, and operation. Real-time operational mode analysis promptly identifies anomalies and potential risks in the distribution network, providing timely and accurate operational information to operations and maintenance personnel, enabling rapid response and decision-making.

[0048] S102, dynamically calculating the protection range sensitivity according to the topology structure and real-time operation mode, and based on the protection range sensitivity and electrical quantity information, using a preset three-stage action rule to perform detection and judgment to obtain a real-time protection action signal.

[0049] In this embodiment, a three-stage protection device is designed using a microprocessor and corresponding protection algorithms. The setting values ​​for the current quick-trip protection, time-limited current quick-trip protection, and overcurrent protection are appropriately determined based on the distribution network's operating parameters and short-circuit current calculations. When the collected current exceeds the corresponding setting value, the protection device issues an action signal within the set time limit.

[0050] Specifically, the sensitivity of each protection zone is dynamically calculated based on the topology model and real-time operating conditions (such as load conditions and power supply status). Sensitivity is typically expressed as the ratio of the short-circuit current at the end of the protection zone to the circuit breaker's set current, ensuring timely circuit breaker operation when a fault occurs.

[0051] Stage 1 (current quick-trip protection): The set point is set relatively high, with no set-up time delay. Once the detected current exceeds this set point, the protection device will immediately operate, rapidly disconnecting the faulty circuit. This is primarily used to protect against short-circuit faults in the vicinity of a line. It operates quickly, but its protection range is limited.

[0052] Stage 2 (Time-Limited Current Instantaneous Trip): This protection has a lower setting value than the instantaneous trip protection and introduces a setting time delay. This protection operates only after the line current reaches the set value and persists for a period of time. The setting must ensure that it covers the entire length of the line and extends appropriately to the front half of the next-level line. This protection serves as the primary protection for the line and also provides remote backup protection for the next-level line.

[0053] The third stage (definite time overcurrent protection) has a lower setting value and a longer time delay than the previous two. It not only ensures full line coverage but also offers a longer protection range than the time-limited current instantaneous trip protection. As backup protection for the line, it also provides remote backup protection for the next lower level and even lower levels.

[0054] Based on the sensitivity of the protection range and real-time collected electrical quantity information, the distribution network is monitored in real time. The real-time monitored data is matched with the preset three-stage action rules. Once a match is successful, a fault is detected and the corresponding protection action signal is immediately triggered, controlling the circuit breaker to operate and disconnect the faulty circuit.

[0055] In this embodiment, by dynamically calculating the sensitivity of the protection range, it is possible to more accurately determine whether a fault occurs within the protection area, thereby avoiding false operation or refusal to operate. By adopting a three-stage action rule, it is possible to quickly respond to different fault conditions, promptly cut off the fault circuit, and reduce the impact of the fault on the power grid. Through real-time monitoring and rapid protection action, it is possible to effectively prevent the expansion and spread of faults, ensure the stable operation of the power grid, improve the power supply reliability of the power grid, reduce the duration and scope of power outages, and enhance user satisfaction. By dynamically adjusting the protection range and sensitivity according to the topology structure and real-time operation mode, it is possible to optimize the protection configuration and reduce unnecessary investment in protection equipment. This reduces operation and maintenance costs and improves the economic benefits of the power grid.

[0056] S103, collect the action timing data of each protection device of the distribution network in real time, build an FA fault feature matrix based on the action timing data, electrical quantity information and topological structure, and use the LSTM neural network to analyze the time series correlation of the FA fault feature matrix to generate an FA fault probability distribution map.

[0057] In this embodiment, the FA system's monitoring terminals collect real-time voltage and current information from each monitoring point and transmit this data to a master station system via a communications network. The master station system builds a fault analysis model based on the distribution network's topology and electrical parameters. When a fault occurs, the master station compares changes in electrical quantities at different monitoring points, such as voltage dips and current surges, to determine whether a fault exists and uses a fault location algorithm to initially determine the approximate fault area.

[0058] Specifically, in the distribution network, a data acquisition module is configured for each protection device to collect its action sequence data in real time, including action time and action type (e.g., tripping, closing). The protection device action sequence data, electrical quantity information, and topology data are integrated to form a dataset containing multidimensional features. Features relevant to fault analysis are extracted from this integrated dataset, such as the interval between action sequences, the amplitude of electrical quantity changes, and the connectivity of the topology. The extracted features are arranged in time series to construct a fault feature matrix. Each row of this matrix represents a feature vector at a time point, and each column represents a feature dimension.

[0059] Build an LSTM neural network model capable of processing time series data and capturing temporal correlations within the feature matrix. Train the LSTM model using historical fault data and adjust its parameters to accurately identify fault features and predict fault probabilities. Input the FA fault feature matrix into the trained LSTM model, which analyzes the temporal correlations within the feature matrix to capture the precursors and patterns of fault occurrence. The LSTM model outputs the probability of failure at each time point, reflecting the likelihood of failure occurring at different time points. Visualize the calculated failure probabilities as a time series to generate a FA failure probability distribution graph. This graph visually illustrates the changing trend of failure probabilities over time.

[0060] In this embodiment, by collecting real-time protection device action time series data, electrical quantity information, and topology information, an FA fault feature matrix is ​​constructed, which can more comprehensively reflect the grid's operating status and fault characteristics. Using an LSTM neural network to analyze time series correlations can capture the precursors and patterns of fault occurrence, improving the accuracy of fault prediction. Generating an FA fault probability distribution map can intuitively reflect the changing trend of fault probability over time, providing operators with intuitive fault warning information. Based on the fault probability distribution map, operators can formulate more reasonable maintenance plans and emergency measures, thereby improving the reliability and safety of the grid.

[0061] S104 , based on the real-time protection action signal and the FA fault probability distribution map, a fault location algorithm is used to perform collaborative analysis to obtain a fault location result.

[0062] In this embodiment, the collaborative control unit utilizes a high-performance computer and data processing algorithms to receive the actuation signals of the three-stage protection device and the fault analysis results from the FA system. First, based on the actuation range of the three-stage protection device and the fault zone determined by the FA system, the system eliminates areas where the fault is unlikely to exist, thereby narrowing the fault range. Then, using a fault location algorithm based on electrical quantity characteristics, the system accurately calculates the narrowed fault zone and locates the fault point.

[0063] Specifically, select an algorithm suitable for distribution network fault location, such as an impedance-based fault location algorithm or a traveling wave-based fault location algorithm. For example, the impedance method uses the voltage and current measured at the time of the fault to calculate the impedance of the fault circuit. The fault location is then determined based on the line impedance parameters.

[0064] The real-time protection action signal is integrated with the FA fault probability distribution map. For example, the protection device location information in the protection action signal is matched with the location information in the fault probability distribution map to determine the fault probability corresponding to each protection action.

[0065] Based on the protection action signal, the fault zone is determined. For example, if a circuit breaker trips, the fault may have occurred in the line zone protected by the circuit breaker.

[0066] Within the initially determined fault range, the fault probability information provided by the FA fault probability distribution map is used to further narrow the fault location. The location with the highest fault probability is selected as the final fault location result.

[0067] Compare and verify the fault location results with the actual fault location. The actual fault location can be obtained through on-site inspections, equipment testing, and other methods. If the location results match the actual fault location, the fault location algorithm is effective. If not, the algorithm needs to be optimized and adjusted. The fault location results are output in an intuitive manner, such as by marking the fault location on a map or generating a fault report. The results are also transmitted to relevant operations and maintenance personnel so they can take timely repair measures.

[0068] In this embodiment, by combining real-time protection action signals with the FA fault probability distribution map, it is possible to fully utilize the protection device action information and fault probability information to more accurately determine the fault location. Compared with traditional single-fault location methods, this method considers more factors and reduces location errors. Accurate fault location results can help operation and maintenance personnel quickly find the fault point, reducing troubleshooting time. Based on the location results, operation and maintenance personnel can directly go to the fault site for repair, improving fault repair efficiency, shortening power outage duration, and reducing power outage losses. Timely and accurate fault location helps quickly restore normal operation of the power grid, reduces the impact of the fault on the grid, improves the reliability and stability of the grid, and guarantees users' electricity needs.

[0069] S105 , based on the fault location results, a reinforcement learning algorithm is used to dynamically plan the load transfer path of the distribution network to generate an optimal isolation and restoration plan.

[0070] In this embodiment, the collaborative control unit issues control instructions based on the fault location results, controlling the corresponding switchgear. For example, if the fault is located on a branch line, the circuit breaker on that branch line is tripped, isolating the faulted area. Simultaneously, the FA system automatically selects and closes the appropriate tie switch based on the distribution network topology and load conditions, transferring the load in the non-faulty area to other power sources and restoring power. A backup battery power supply ensures independent local protection in the event of a power outage.

[0071] Specifically, the distribution network's topology, load distribution, and switch status are used as the state space for reinforcement learning. Each state represents the operating state of the distribution network at a specific moment. Switching operations in the distribution network (such as closing and tripping) are used as the action space for reinforcement learning. Each action represents a change to the distribution network's topology. A reward function is designed to encourage the algorithm to find the optimal load transfer path. The reward function can include multiple indicators such as load recovery, line loss reduction, and the number of switch operations. For example, positive rewards are given when load recovery increases, line losses decrease, or the number of switch operations decreases; negative rewards are given when these decreases.

[0072] Deep reinforcement learning algorithms (such as DQN and DDPG) are trained to solve the load transfer problem in distribution networks. During training, the algorithm learns to find the optimal load transfer path by repeatedly trying different switching combinations. Based on the trained reinforcement learning model, the algorithm dynamically plans the load transfer path for the distribution network based on real-time electrical quantity information and fault location results. The algorithm selects the optimal action (i.e., switching operation) based on the current state (i.e., the operating status of the distribution network) to quickly restore loads and minimize line losses. Based on the fault location results and the load transfer path generated by the reinforcement learning algorithm, an optimal isolation plan is developed. This isolation plan includes disconnecting the faulty line and isolating the faulty area to prevent the fault from expanding and spreading. After isolating the faulty area, loads in non-faulty areas are transferred to other reliable power lines based on the load transfer path. The restoration plan should include closing the circuit breaker and distributing the load to ensure that loads in non-faulty areas are restored to power as quickly as possible.

[0073] In this embodiment, the reinforcement learning algorithm can automatically learn the operating patterns and load characteristics of the distribution network and find the optimal load transfer path, thereby improving the efficiency and accuracy of load transfer and reducing the risk of manual intervention and misoperation. By dynamically planning the load transfer path and generating the optimal isolation and restoration plan, power supply to non-faulty areas can be quickly restored after a fault occurs, reducing the duration and scope of power outages and improving the reliability of the power grid. The reinforcement learning algorithm considers line loss factors when planning load transfer paths, and strives to select paths with lower line loss for transfer, thereby reducing the line loss rate of the power grid and improving the economic benefits of the power grid. At the same time, the automated load transfer process reduces the workload of operation and maintenance personnel and reduces operation and maintenance costs.

[0074] In some embodiments, in the above step S101, obtaining the topology and real-time operation mode of the distribution network by preprocessing and extracting features from the electrical quantity information specifically includes:

[0075] The electrical quantity information is decomposed and reconstructed using a wavelet threshold denoising algorithm, the electrical quantity information is data compensated using a Lagrange interpolation method, and the decomposed, reconstructed and data compensated electrical quantity information of each monitoring point is time-aligned to obtain standard electrical quantity information;

[0076] Constructing a voltage and current video feature matrix based on the standard electrical quantity information, performing node correlation calculation on the voltage and current video feature matrix, and generating a multidimensional feature vector set representing the spatiotemporal correlation of electrical quantities;

[0077] Using the node association threshold and switch state quantity in the multidimensional feature vector set, dynamically generating and updating the adjacency matrix, and reconstructing the topological graph model of the distribution network;

[0078] Based on the topology model, through distributed power flow calculation and overload index analysis, the operation status mapping table and abnormal warning signal of the distribution network are output to obtain the real-time operation mode of the distribution network.

[0079] In this embodiment, a wavelet threshold denoising algorithm is used to decompose and reconstruct electrical quantity information. Lagrange interpolation is then used to compensate for the decomposed and reconstructed electrical quantity information, generating preliminary processed electrical quantity information. Time alignment is performed on the preliminary processed electrical quantity information at each monitoring point to obtain standard electrical quantity information. A voltage and current video feature matrix is ​​constructed based on the standard electrical quantity information. The numerical distribution characteristics of each element in the matrix are calculated to obtain feature matrix distribution data. Node associations are calculated for the feature matrix distribution data to generate a set of associations representing the relationships between nodes. Multidimensional features are extracted from the set of associations to generate a set of multidimensional feature vectors representing the spatiotemporal correlations of electrical quantities.

[0080] By analyzing node associations and switch states, we obtain correlation data, dynamically update the adjacency matrix, and obtain reconstructed matrix data. Using this reconstructed matrix data, we generate a distribution network topology model and determine the network structure. We then perform distributed power flow calculations on the topology model to obtain power flow distribution data. Using this power flow distribution data, we calculate overload indicators and determine overload areas. If the overload indicator exceeds the threshold setting, we generate an abnormal warning signal and determine the warning area. We then obtain operating status data and, combined with the abnormal warning signal, generate an operating status mapping table to determine the real-time operating mode. We then extract change trends from the real-time operating mode, update the threshold settings and matrix update strategy, and obtain optimized operating parameters.

[0081] For example, assuming the voltage signal at a distribution network monitoring point is affected by high-frequency noise, a wavelet transform can be used to decompose the signal into different frequency components. A threshold is then set to filter out the noise, and a smoothed signal can be reconstructed. This method can effectively improve data quality and provide a reliable foundation for subsequent analysis. In one possible implementation, if the original voltage signal amplitude is 220V and the added noise fluctuates to 215-225V, wavelet denoising can restore it to a stable value close to 220V, significantly reducing errors. Using Lagrange interpolation to compensate for electrical quantity information primarily addresses issues such as missing data or uneven sampling. Specifically, if current data at a monitoring point is missing at a certain moment due to equipment failure, Lagrange interpolation can be used to estimate the value at time t2 to approximately 11A based on known data points before and after, such as 10A at time t1 and 12A at time t3. This compensation method is simple and efficient, ensuring data continuity. Preferably, if there are many missing points in actual business scenarios, time window smoothing can be introduced to further improve compensation accuracy. In one possible implementation, time alignment of the decomposed, reconstructed, and compensated electrical quantity information can be achieved through timestamp calibration. For example, due to different sampling frequencies, data timestamps at multiple monitoring points in a distribution network may be misaligned. For example, data at point A may be at 10:00:00, while data at point B may be at 10:00:02. By aligning the data to a unified time axis of 10:00:00, standard electrical quantity information is generated. This alignment lays the foundation for subsequent matrix construction, preventing time deviations from affecting analysis results. When constructing a voltage and current video feature matrix based on standard electrical quantity information, the temporal variations of voltage and current can be considered as "video frames." For example, a distribution network has three nodes, each recording voltage and current data at five time instants. A 3×10 matrix can be constructed, with rows representing nodes and columns representing voltage and current sequences. Node correlation is calculated on this matrix, for example, by using correlation coefficients to analyze the spatiotemporal correlation of electrical quantities between nodes, generating a multidimensional feature vector set. This method intuitively reflects the dynamic characteristics of the power grid and helps detect abnormal patterns. Dynamically generating and updating the adjacency matrix using node association thresholds and switch status values ​​from a multidimensional feature vector set can be understood as reshaping the topological relationship based on electrical connections and switch states. For example, if a node's association falls below the threshold of 0.8 and the switch is disconnected, the corresponding element in the adjacency matrix is ​​set to 0; otherwise, it is set to 1. This dynamic update can reflect changes in the distribution network topology, such as line disconnections or grid-connected operations, in real time, improving model adaptability. When performing distributed power flow calculations and overload index analysis based on the topological graph model, the operating status can be determined by the power flow direction of each node. In one possible implementation, assuming a line has a power flow of 500kW, a rated capacity of 600kW, and an overload index of 83%, which is below the warning threshold of 90%, a normal state mapping table is output. If the power flow of another line reaches 650kW and the overload index exceeds the standard, an abnormal warning signal is triggered. This analysis method can quickly locate problem areas and ensure the safe operation of the power grid.It should be noted that the above method forms a complete technical chain through denoising, compensation, alignment, feature extraction, topology reconstruction and state analysis. Each step improves data accuracy and analysis reliability, and ultimately realizes accurate monitoring and optimized management of the real-time operation mode of the distribution network.

[0082] In some embodiments, in the above step S102, the protection range sensitivity is dynamically calculated according to the topology structure and the real-time operation mode, and a preset three-stage action rule is used to perform detection and judgment based on the protection range sensitivity and electrical quantity information to obtain a real-time protection action signal, specifically including:

[0083] Based on the topology and real-time operation mode, a directional sensitivity coefficient is set for each branch of the topology, wherein the sensitivity coefficient is used to characterize the response strength of the branch to downstream faults;

[0084] Constructing a sensitivity matrix based on the directional sensitivity coefficients, performing a Hadamard product operation on the sensitivity matrix and the adjacency matrix, eliminating invalid branches, and obtaining an optimized sensitivity matrix, wherein the sensitivity matrix is ​​used to determine the protection range sensitivity of each branch;

[0085] Dynamically generate a three-stage protection criterion set with adaptive threshold and time delay characteristics based on the sensitivity matrix and the real-time load current in the electrical quantity information;

[0086] By performing topological connectivity verification on the three-stage protection criterion set, a real-time protection action signal for the fault branch is generated.

[0087] In this embodiment, the directional sensitivity coefficient of each branch is obtained based on the topology and operating mode. The coefficient represents the response strength. A preset threshold is used to determine the validity of the branch direction to obtain an initial sensitivity set. A sensitivity matrix is ​​constructed from the initial sensitivity set. Based on the correspondence between the branch direction and the topology, the distribution characteristics of the response strength are determined by assigning values ​​to the matrix elements. An adjacency matrix is ​​obtained, and the sensitivity matrix and the adjacency matrix are processed using a Hadamard product operation. Invalid branches are eliminated to obtain an optimized first matrix. The protection range is analyzed using the first matrix. If the response strength of a branch is lower than the preset threshold, it is determined to be an invalid range, and the optimized matrix is ​​updated to the second matrix. The second matrix is ​​used to classify the branches using the K-means clustering algorithm. The protection range boundaries of each branch are determined based on the classification results to obtain a range allocation matrix. The priority sequence of the fault response is calculated using the range allocation matrix, and the optimization matrix is ​​adjusted according to the priority sequence to obtain the final sensitivity distribution. The protection range sensitivity of each branch is extracted from the final sensitivity distribution, and a linear regression algorithm is used to predict the changing trend of the fault response to obtain a dynamic adjustment strategy for branch protection.

[0088] By combining the sensitivity matrix with real-time load calculations to dynamically generate adaptive thresholds and delay characteristics, a three-stage protection criterion set is constructed based on the adaptive thresholds and delay characteristics. By integrating the criterion set, the data structure required for topology connectivity verification is obtained. If the topology connectivity verification passes, the faulty branch location is determined. Based on the faulty branch location, the corresponding protection action instructions are generated. The protection action instructions generate real-time signals for the faulty branch. By comparing the real-time signals with the current information, the system's operating status is determined.

[0089] For example, when setting a directional sensitivity coefficient for each branch in a distribution network topology, it can be understood as a metric that measures the branch's impact on downstream faults. For example, suppose a branch connects nodes A and B. If a short circuit occurs at downstream node B, the current direction and amplitude in the branch will change significantly. The directional sensitivity coefficient can be determined by analyzing historical fault data and branch location. For example, a branch near a load center might be assigned a higher coefficient, such as 0.9, while a remote branch might be assigned a lower coefficient, such as 0.5. This design more accurately reflects the branch's response to faults. In one possible implementation, when constructing a sensitivity matrix based on the directional sensitivity coefficients, the distribution network can be simplified as a three-node system. Assume that the branch sensitivity coefficients from node 1 to node 2 are 0.8, and from node 2 to node 3 are 0.6. The matrix is ​​initially constructed in a 3×3 format. By performing a Hadamard product with the adjacency matrix, invalid branches are removed. For example, if a branch has no electrical connection due to a disconnected switch, the corresponding element is set to 0. The optimized sensitivity matrix clearly shows the sensitivity of the protection range of each branch. For example, the branch protection range from node 1 to node 2 is more inclined to respond quickly to downstream large current faults. Specifically, when a three-stage protection criterion set is generated based on the sensitivity matrix and real-time load current, it can be dynamically adjusted in combination with actual operating data. Assuming that the normal load current of a branch is 100A, the sensitivity matrix shows that its coefficient is 0.7, and the current suddenly increases to 500A under a short-circuit fault. The criterion set can be set as: the first stage instantaneous action at 400A, the second stage with a delay of 0.5 seconds to act at 300A, and the third stage with a long delay of 1 second to act at 200A. This adaptive threshold and time delay characteristic can flexibly respond to different fault scenarios and avoid false or missed actions. Preferably, when integrating the three-stage protection criterion set and performing topological connectivity verification, the accuracy of the protection action can be ensured by checking the connectivity status of the fault branch with the upstream and downstream nodes. For example, if a node downstream of a branch loses power due to a line disconnection, the validation criteria set will eliminate protection action signals for that node and generate real-time protection commands only for branches that maintain electrical connectivity. This approach effectively reduces redundant operations and improves system response efficiency. In one possible implementation, the generation of real-time protection action signals for the faulty branch can further incorporate the specific location of the fault. Suppose a ground fault occurs in the branch downstream of node 3 in a distribution network. The sensitivity matrix shows a coefficient of 0.6, and the real-time current jumps to 450A, exceeding the threshold of the first-stage criteria. The system immediately generates a protection action signal to cut off power to that branch. This fast-response signal generation method can promptly isolate the faulty area and ensure normal operation of other branches. It should be noted that the introduction of directional sensitivity coefficients makes the protection scope more targeted. For example, if directional sensitivity is not set and judgment is based solely on current amplitude, it may lead to false action on upstream branches. However, by optimizing the sensitivity matrix, the fault impact area can be clearly identified, reducing unnecessary power outages. This technical approach significantly improves the accuracy and reliability of distribution network protection.As you can see, the adaptive nature of the three-stage protection criteria set provides flexibility for varying operating conditions. During peak load periods, the threshold can be adjusted upward to accommodate normal fluctuations; during off-peak load periods, the threshold can be lowered to quickly detect anomalies. This dynamic adjustment ensures that the protection strategy closely matches the grid's operating status, providing strong support for real-time monitoring.

[0090] Furthermore, the three-stage protection criterion set specifically includes:

[0091]

[0092]

[0093]

[0094]

[0095] in, Indicates the action current threshold of section I, Indicates the action current threshold of section II, Indicates the action current threshold of section III, Represents the reliability factor used to ensure that the operating current threshold of stage I is higher than the maximum short-circuit current. Indicates the coordination coefficient with the I-stage action current threshold, Indicates the overload protection factor, represents the sensitivity coefficient, Indicates the maximum three-phase short-circuit current at the end of the branch. represents the time step constant, represents the sensitivity and time adjustment factor, Indicates the first stage of action time, Indicates the action time of the second stage, Indicates the action time of section III, Indicates the real-time load current of the branch. represents the time margin, Indicates the effective value of the current measured by the protection device in real time. represents the rate of change of current, Indicates the duration of the current exceeding the threshold.

[0096] In this embodiment, For the first section (instantaneous quick break), For Section II (time-limited quick disconnection), This is Section III (time-limited quick disconnection).

[0097] The first stage operating current threshold is used for quick-trip current protection and is set high to ensure instantaneous operation and rapid interruption of the fault current in the event of a severe short-circuit fault in the vicinity of the line. Its value is typically set based on the maximum short-circuit current that could occur at the next line exit under the system's maximum operating mode, multiplied by a reliability factor to ensure selectivity.

[0098] The operating current threshold for Section II is used for time-limited current instantaneous trip protection. Its value is lower than that of Section I but higher than that of Section III. It is intended to protect the entire length of the line and serve as a backup for the instantaneous current trip protection. Its setting must be coordinated with the instantaneous current trip protection of the next section of the line, typically using a coordination factor.

[0099] The third stage action current threshold is used for time-limited overcurrent protection. It has the lowest setting value and is adjusted according to the maximum load current to be avoided. It is multiplied by the overload protection coefficient to ensure that there will be no false operation during normal operation.

[0100] The reliability coefficient is used to ensure that the operating current threshold of stage I is higher than the maximum short-circuit current, so that it can operate reliably even in the maximum operating mode of the system, thereby improving the reliability of protection.

[0101] The coordination coefficient is used to coordinate the operating current thresholds between adjacent protections to ensure that only the protection closest to the fault point is activated during a fault, thus achieving protection selectivity.

[0102] The overload protection factor is used in the time-limited overcurrent protection to ensure that the operating current threshold is higher than the maximum load current to prevent false operation during normal load fluctuations.

[0103] The sensitivity coefficient is used to measure the sensitivity of the protection to faults. It is usually related to the ratio of the minimum short-circuit current to the operating current threshold, ensuring that the protection can operate reliably even in minor faults.

[0104] The maximum three-phase short-circuit current at the end of the branch is used to set the operating current threshold, especially in current quick-break protection, to ensure that the protection range covers the important part of the line.

[0105] The time step constant is used to set the action time difference between adjacent protections, ensuring that only the protection closest to the fault point is activated first in the event of a fault, thus achieving time selectivity.

[0106] The sensitivity and time adjustment factors are used to adjust the action time according to the sensitivity requirements to ensure timely action even at the edge of the protection range or in the event of a minor fault.

[0107] Action time 、 and They correspond to the action time of Section I, Section II and Section III protection respectively, ensuring that they act in the set sequence and time intervals in case of a fault.

[0108] The branch real-time load current is used to monitor the line load in real time and is compared with the action current threshold to determine whether action is required.

[0109] The time margin is used to add a certain margin to the action time to cope with factors such as measurement errors and system fluctuations to ensure the reliability of protection.

[0110] The effective current value measured by the protection device in real time is used to compare with the action current threshold in real time to determine whether protection needs to be started.

[0111] The current change rate is used in some protection systems to assist in determining the type or severity of a fault and to improve the sensitivity and speed of protection.

[0112] The current exceeding threshold duration is used in the definite time overcurrent protection. When the current exceeds the operating current threshold and the duration reaches the set value, the protection will be activated to prevent false operation caused by instantaneous overcurrent.

[0113] In this embodiment, by reasonably setting the action current threshold, coordination coefficient and time step constant, it is ensured that only the protection closest to the fault point is actuated in the event of a fault, thereby avoiding over-tripping and expanding the scope of power outage.

[0114] Enhanced protection speed: Current quick-trip protection operates instantaneously, quickly clearing severe short-circuit faults and minimizing equipment damage and power outages. By adjusting parameters such as sensitivity and the rate of change of monitored current, protection can be reliably activated even for minor faults, improving grid safety and reliability. Time-limited current quick-trip protection and time-limited overcurrent protection can protect the entire length of the line, ensuring timely and effective protection regardless of fault location. By dynamically adjusting operating current thresholds and time parameters, protection can adapt to changes in system operation, enhancing its adaptability and flexibility.

[0115] In some embodiments, in the above step S103, the FA fault feature matrix is ​​constructed based on the action timing data, electrical quantity information, and topological structure, and the LSTM neural network is used to analyze the timing correlation of the FA fault feature matrix to generate the FA fault probability distribution map, which specifically includes:

[0116] Align and tensor-join the action timing data, electrical quantity information, and topological structure to form the FA fault feature matrix;

[0117] Construct a two-layer bidirectional LSTM network, input the FA fault feature matrix into the two-layer bidirectional LSTM network for time series correlation analysis, and obtain the fault probability value of each coordinate of the distribution network;

[0118] The fault probability values ​​of each coordinate of the distribution network are output in the form of a heat map to generate a FA fault probability distribution map.

[0119] In this embodiment, raw data is acquired through action timing, electrical quantity signals, and topological structure. Data alignment methods are used to ensure consistency among these three data sets, resulting in an aligned dataset. Tensor splicing technology is used to fuse the aligned datasets and generate an FA fault feature matrix. A two-layer bidirectional LSTM network is constructed and the FA fault feature matrix is ​​input into the network for training, resulting in a time series association model. The FA fault feature matrix is ​​analyzed using the time series association model to obtain the fault probability of each distribution network station. If the fault probability exceeds a preset threshold, the corresponding distribution network station is deemed to have a fault risk, resulting in a set of fault coordinates. Based on the set of fault coordinates, a coordinate analysis method is used to determine the fault area distribution, resulting in a regional fault distribution. This regional fault distribution is used as input to generate a graphical representation of the fault probability using a heat map algorithm, resulting in a preliminary distribution map. If abnormal areas are found in the preliminary distribution map, the graphical parameters are adjusted by comparing the probability values ​​to obtain an optimized distribution map. Key analysis results are extracted from the optimized distribution map to determine the FA fault probability distribution characteristics. A final FA fault probability distribution map is obtained and output as a heat map.

[0120] For example, the alignment and tensor concatenation of action time series data, electrical quantity information, and topology can be understood as a multi-dimensional data fusion process. Action time series data may include time series of switch trips, electrical quantity information includes real-time current and voltage values, and the topology reflects the node connectivity of the distribution network. Suppose a distribution network has three nodes. At 10:00 a.m., a switch action occurs on the branch from node 1 to node 2, causing a sudden increase in current from 100A to 300A. The topology indicates a unidirectional power supply path. After alignment, the data is organized into a three-dimensional tensor, for example, with a time dimension of 10-second samples, an electrical quantity dimension including current and voltage, and a topology dimension indicating branch status. This concatenation method fully preserves the temporal and spatial characteristics of the data, providing a comprehensive foundation for subsequent analysis. In one possible implementation, a two-layer bidirectional LSTM network can be constructed as a tool for capturing temporal dependencies. The two-layer design enhances the model's ability to extract complex patterns, while the bidirectional design considers both past and future contextual information. Specifically, suppose the input FA fault feature matrix contains 5 consecutive minutes of current data for a branch, with values ​​of 100A, 120A, 150A, 300A, and 200A, respectively. A two-layer bidirectional LSTM network analyzes the trend of this data. For example, a sudden increase from 150A to 300A may indicate a fault, while a subsequent drop to 200A may indicate fault mitigation. Through multi-directional temporal correlation analysis, the network generates a fault probability value for each node, such as 0.85 for node 2 and 0.25 for node 3. This approach fully leverages the dynamic nature of the data. It should be noted that when the fault probability values ​​are output as a heat map, the heat map can intuitively reflect the fault distribution of the distribution network. For example, a probability of 0.85 for node 2 may appear as a red high-risk area, while a probability of 0.25 for node 3 may appear as a green low-risk area. In one possible implementation, assuming a distribution network covering 10 nodes, the heat map shows that nodes near the load center generally have probabilities above 0.7, while edge nodes have probabilities below 0.4. This visualization method makes it easier for operators to quickly locate high-risk areas and improve decision-making efficiency. Preferably, when generating the FA fault probability distribution map, it can be further refined in combination with actual scenarios. Suppose that the insulation of a branch has decreased due to aging, and the current data has fluctuated many times in the past hour. After LSTM network analysis, a failure probability of 0.9 is given. In the heat map, the branch is marked in dark red, indicating that it needs priority maintenance. It can be understood that this distribution map not only reflects the current status, but also provides a basis for preventive maintenance. For example, if the probability of a certain area is high for many consecutive days, equipment replacement can be arranged in advance to reduce the occurrence rate of failures. In one possible implementation method, the reliability of time series correlation analysis can be verified from different aspects. Suppose that the current of a node suddenly increases to 250A when the load is low in the early morning, but there is no abnormality during the peak in the daytime.By analyzing the preceding and following time series, a two-layer bidirectional LSTM identifies a sudden increase during a valley as a fault sign with a probability of 0.8, while the probability of normal fluctuation during a peak period is only 0.2. This multi-period comparison enhances the accuracy of the analysis. If the topology indicates a branch downstream of the node, the network further identifies the fault location based on the branch current distribution. This multi-dimensional support ensures the credibility of the probability values ​​and provides strong technical support for distribution network management.

[0121] In some embodiments, in step S104, the method of performing collaborative analysis based on the real-time protection action signal and the FA fault probability distribution map using a fault location algorithm to obtain a fault location result specifically includes:

[0122] Based on the real-time protection action signal and FA fault probability distribution map, an improved matrix algorithm is used to analyze and generate a candidate set of fault areas.

[0123] Based on the fault area candidate set, transient current features are extracted using wavelet packet transform to construct a decision tree classifier;

[0124] Based on the fault area candidate set, a protection action logic is temporally reasoned using a Petri net to obtain a protection action logic rule base;

[0125] A collaborative analysis is performed based on the decision tree classifier and the protection action logic rule base to generate a fault location result.

[0126] In this embodiment, features are extracted from the fault probability values ​​and processed using a matrix algorithm to generate preliminary analysis results. If the preliminary analysis results exceed a preset threshold, the matrix algorithm is optimized using an improved method to obtain an adjusted result. Based on the adjusted results and the distribution diagram, fault analysis techniques are used to determine the candidate region range. The candidate region range is processed using a candidate generation technique to obtain a candidate set of fault regions. The candidate set of fault regions is obtained and grouped using a clustering algorithm to determine the final fault region. Using the candidate set of fault regions, transient current signals are decomposed using a wavelet packet transform to obtain a feature set. Based on the protection action data, temporal reasoning is modeled using Petri nets to obtain a logical rule base. A decision tree classifier is constructed based on the feature set to generate preliminary classification results. Constraint verification is performed on the preliminary classification results using the logical rule base to obtain adjusted classification data. If the adjusted classification data matches the fault region, the fault location result is determined. If not, the feature set is updated through collaborative analysis. The decision tree classifier is rerun using the updated feature set to obtain an optimized classification result. The optimized classification result is integrated with the logical rule base to determine the final fault location result.

[0127] For example, an improved matrix algorithm can be used to generate a candidate set of fault regions based on real-time protection action signals and a fault probability distribution map for fault management (FA). The protection action signal may include the timestamp of the switch tripping, while the FA fault probability distribution map provides the fault probability for each node. Suppose a distribution network has five nodes. Node 3 experiences a protection action at 10:05. The probability distribution map shows a probability of 0.9, while the probability for neighboring node 4 is 0.6. The improved matrix algorithm organizes this information into a matrix and, combining the signals and probabilities, identifies high-risk areas, such as nodes 3 and 4, to form a candidate set. This approach can quickly narrow the fault scope. In one possible implementation, using wavelet packet transform to extract transient current features can be considered a tool for decomposing current signals. Suppose the current at node 3 suddenly changes from 150A to 350A at the moment of tripping. The wavelet packet transform decomposes this change into high-frequency and low-frequency components, with the high-frequency component reflecting the sudden change. By analyzing these features, a decision tree classifier is constructed. For example, when the high-frequency component exceeds a certain threshold, the classifier identifies a fault signal. This approach effectively distinguishes normal fluctuations from transient faults. Specifically, using Petri nets to perform sequential reasoning on protection action logic based on a candidate set of fault regions can be understood as a dynamic modeling process. Suppose that after node 3 trips, the relay at node 4 operates at 10:06. The Petri net, through state transitions, infers whether the action sequence complies with the pre-set logic, such as "After node 3 trips, node 4 should respond with a delayed response." This ultimately generates a rule base, such as "When node 3 fails, node 4 must be isolated." This provides a logical basis for subsequent analysis. It should be noted that the decision tree classifier and the protection action logic rule base can complement each other when performing collaborative analysis. Suppose the decision tree determines node 3 as the fault source based on current characteristics with a probability of 0.95, while the rule base verifies that the action of downstream node 4 complies with logic. After collaborative analysis, node 3 is confirmed to be the fault point. This combination improves location reliability. Preferably, in one possible implementation, assume that the distribution network covers eight nodes. A short circuit causes the current at node 5 to suddenly increase to 400A. After extracting features using wavelet packet transform, the decision tree determines a fault probability of 0.88. Petri net reasoning revealed that after node 5 tripped, node 6 was not promptly isolated, prompting the rule base to indicate an anomaly. After collaborative analysis, the fault was located at node 5, and a recommendation was made to inspect the protection device at node 6. This approach not only locates the fault but also identifies potential issues. As can be appreciated, the analysis process, from candidate set generation to final location, proceeds logically in a progressive manner. For example, a matrix algorithm first defines the scope, wavelet packet transforms and decision trees focus on features, and Petri nets verify the logic, ultimately leading to a collaborative output of the results. Suppose a branch circuit experiences abnormal current due to a lightning strike. The candidate set contains three nodes. After feature extraction, node 2 is identified. Logical reasoning confirms the rationality of its protective action, ultimately locating node 2 as the fault point. This multi-step collaboration ensures the credibility of the results.In one possible implementation, consider a node whose current suddenly surges to 300A in the early morning and returns to normal during the day. A matrix algorithm identifies this node as a candidate set, a wavelet packet transform extracts the mutation signature, a decision tree identifies it as a fault, and a Petri net combines upstream and downstream action verification logic to ultimately locate the node. This multi-faceted analysis enhances accuracy and provides operators with clear fault location information.

[0128] In some embodiments, in step S105, dynamically planning the load transfer path of the distribution network using a reinforcement learning algorithm based on the fault location result to generate an optimal isolation and restoration plan specifically includes:

[0129] The fault location result is used as the state space, the switch operation instruction set is used as the action space, and minimizing the switch cost and load loss is used as the reward function;

[0130] Based on the state space, action space and reward function, the DQN reinforcement learning algorithm is used to make action decisions and output the optimal action sequence;

[0131] According to the optimal action sequence, the switch operation sequence table and the corresponding load transfer path are analyzed to generate the optimal isolation and restoration plan.

[0132] In this embodiment, a state space is constructed based on the fault location results to obtain a set of all possible states. A set of switch operation instructions is generated from the state space to form an action space, which determines the action range. A reward function is calculated by minimizing switching cost and load loss to obtain a basis for action evaluation. Based on the state space, action space, and reward function, a DQN algorithm is applied to make action decisions and output the optimal action sequence. The optimal action sequence is used to analyze the switch operation sequence table and determine the load transfer path. Based on the load transfer path, an optimal isolation plan is generated to determine the fault isolation steps. The recovery plan is adjusted based on the isolation plan and the action sequence to obtain the load recovery path.

[0133] For example, using the fault location results as a state space can be understood as using the identified fault points and their associated state information in the distribution network as the basis for decision-making. For example, if a distribution network has six nodes and node 4 is located as the fault point, the state space would include the fault status of node 4, the switch status of upstream and downstream nodes, and the load distribution. This state space provides a dynamic, global view for subsequent decision-making. In one possible implementation, a set of switch operation instructions serves as the action space, effectively defining all possible operation combinations. Specifically, assuming the distribution network has 10 switches, the action space includes "on" or "off" instructions for each switch, such as "switch S3 open" or "switch S7 close." This set of actions provides the algorithm with an operational scope, ensuring coverage of all possible isolation and restoration paths. It should be noted that minimizing switching cost and load loss as a reward function is crucial for balancing operating costs and power supply reliability. For example, if a single switch operation costs 50 units and the load loss is 100 units per kilowatt-hour, the reward function will favor solutions that require fewer operations and restore more load. Suppose disconnecting S4 isolates the fault but results in a 200 kWh load loss, while disconnecting S5 only results in a 50 kWh load loss. The algorithm prioritizes disconnecting S5. This design makes decisions more aligned with real-world needs. Specifically, the DQN reinforcement learning algorithm can be viewed as an intelligent trial-and-error process when making action decisions. In one possible implementation, assume node 4 fails and the initial load loss is 300 kWh. Through multiple simulations, such as disconnecting S2 and then closing S6, the DQN finds that the load has recovered to 80% and the load loss has been reduced to 60 kWh. The algorithm continuously optimizes and ultimately outputs the optimal action sequence. This approach gradually approaches the optimal solution through learning. Preferably, analyzing the switch operation sequence and load transfer path based on the optimal action sequence converts the decision into an executable plan. For example, if the DQN output sequence is "disconnect S4, close S8," the corresponding sequence table is: the first step is to disconnect S4 to isolate the fault, and the second step is to close S8 to transfer the load to the backup line. The load transfer path may be from node 1 to node 6. This analysis makes the plan intuitive and actionable. It is understandable that the process of generating the optimal isolation and recovery plan progresses step by step from state to action to result. For example, node 5 fails due to a short circuit, and the state space shows that its downstream load is 150 kWh. In the action space, an attempt is made to disconnect S5 and close S9. The reward function evaluation cost is 70 units, and the load loss is reduced to 20 kWh. The final solution is "disconnect S5 and close S9", and the load is transferred to another branch. This method ensures the efficiency of fault isolation and the timeliness of power restoration. In one possible implementation, suppose a node fails in the early morning, and the state space records a load loss of 100 kWh. Through learning, DQN decides to disconnect S3 first and then close S7. The sequence table shows that the two-step operation completes isolation and recovery. The load transfer path turns from node 2 to node 8.This solution not only quickly isolates the fault, but also maximizes load recovery and provides reliable support for operators.

[0134] Reference Figure 2 An embodiment of the present invention provides a system 2 for quickly locating faults based on three-stage protection and FA coordinated protection. The system 2 specifically includes:

[0135] The data acquisition and preprocessing module 201 is used to collect electrical quantity information of each monitoring point in the distribution network in real time, and obtain the topology structure and real-time operation mode of the distribution network by preprocessing and extracting features of the electrical quantity information;

[0136] The three-stage protection module 202 is used to dynamically calculate the protection range sensitivity according to the topology structure and real-time operation mode, and based on the protection range sensitivity and electrical quantity information, use the preset three-stage action rules to perform detection and judgment to obtain a real-time protection action signal;

[0137] The FA fault analysis module 203 is used to collect the action time series data of each protection device in the distribution network in real time, construct an FA fault feature matrix based on the action time series data, electrical quantity information and topological structure, and use the LSTM neural network to analyze the time series correlation of the FA fault feature matrix to generate an FA fault probability distribution map;

[0138] The collaborative control module 204 is configured to perform collaborative analysis using a fault location algorithm based on the real-time protection action signal and the FA fault probability distribution map to obtain a fault location result;

[0139] The fault self-healing module 205 is used to dynamically plan the load transfer path of the distribution network based on the fault location result using a reinforcement learning algorithm to generate an optimal isolation and restoration plan.

[0140] It is understandable that if Figure 1 The contents of the embodiment of the method for quickly locating a fault based on three-stage protection and FA coordinated protection shown in the figure are applicable to the embodiment of the system for quickly locating a fault based on three-stage protection and FA coordinated protection. The functions specifically implemented by the embodiment of the system for quickly locating a fault based on three-stage protection and FA coordinated protection are the same as those in the embodiment of the method for quickly locating a fault based on three-stage protection and FA coordinated protection. Figure 1 The embodiment of the method for quickly locating faults based on three-stage protection and FA coordinated protection is the same as that shown in FIG. Figure 1 The beneficial effects achieved by the embodiment of the method for quickly locating faults based on three-stage protection and FA coordinated protection shown are also the same.

[0141] It should be noted that the information interaction, execution process and other contents between the above-mentioned systems are based on the same concept as the embodiment of the method of the present invention. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0142] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0143] Reference Figure 3 An embodiment of the present invention further provides a computer device 3, comprising: a memory 302, a processor 301, and a computer program 303 stored in the memory 302. When the computer program 303 is executed on the processor 301, a method for quickly locating faults based on three-stage protection and FA collaborative protection as described in any one of the above methods is implemented.

[0144] The computer device 3 may be a desktop computer, a notebook computer, a PDA, a cloud server or other computing devices. The computer device 3 may include, but is not limited to, a processor 301 and a memory 302. Those skilled in the art will understand that Figure 3 This is merely an example of the computer device 3 and does not constitute a limitation on the computer device 3 . The computer device 3 may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the computer device 3 may also include input and output devices, network access devices, etc.

[0145] The processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. A general-purpose processor may be a microprocessor or any conventional processor.

[0146] In some embodiments, the memory 302 may be an internal storage unit of the computer device 3, such as a hard drive or memory of the computer device 3. In other embodiments, the memory 302 may also be an external storage device of the computer device 3, such as a plug-in hard drive, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. equipped on the computer device 3. Furthermore, the memory 302 may include both an internal storage unit of the computer device 3 and an external storage device. The memory 302 is used to store an operating system, application programs, a boot loader, data, and other programs, such as the program code of the computer program. The memory 302 may also be used to temporarily store data that has been output or is about to be output.

[0147] An embodiment of the present invention further provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the method for quickly locating faults based on three-stage protection and FA collaborative protection as described in any one of the above methods is implemented.

[0148] In this embodiment, if the integrated unit is implemented as a software functional unit and sold or used as a standalone product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application can implement all or part of the process steps in the above-mentioned method embodiments by using a computer program to instruct the relevant hardware. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of each of the above-mentioned method embodiments. The computer program includes computer program code, which can be in source code form, object code form, executable file, or some intermediate form. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a camera / terminal device, recording medium, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signals, telecommunication signals, and software distribution media. Examples include USB flash drives, removable hard drives, magnetic disks, or optical disks. In some jurisdictions, based on legislation and patent practice, computer-readable media cannot be electric carrier signals or telecommunication signals.

[0149] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0150] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0151] In the embodiments disclosed in the present application, it should be understood that the disclosed devices / terminal equipment and methods can be implemented in other ways. For example, the device / terminal equipment embodiments described above are merely schematic. For example, the division of the modules or units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0152] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

Claims

1. A method for quickly locating faults based on three-stage protection and FA collaborative protection, characterized in that: The method specifically includes: Real-time collection of electrical quantity information at each monitoring point in the distribution network, and obtaining the topological structure and real-time operation mode of the distribution network by preprocessing and feature extraction of the electrical quantity information; The protection range sensitivity is dynamically calculated based on the topology and real-time operation mode. Based on the protection range sensitivity and electrical quantity information, a preset three-stage action rule is used for detection and judgment to obtain a real-time protection action signal. The system collects the action time series data of each protection device in the distribution network in real time, constructs the FA fault feature matrix based on the action time series data, electrical quantity information and topological structure, and uses the LSTM neural network to analyze the time series correlation of the FA fault feature matrix to generate the FA fault probability distribution map. Based on the real-time protection action signal and FA fault probability distribution map, the fault location algorithm is used for collaborative analysis to obtain the fault location result; Based on the fault location results, a reinforcement learning algorithm is used to dynamically plan the load transfer path of the distribution network and generate the optimal isolation and restoration plan; The method of obtaining the topology and real-time operation mode of the distribution network by preprocessing and extracting features from the electrical quantity information specifically includes: The electrical quantity information is decomposed and reconstructed using a wavelet threshold denoising algorithm, the electrical quantity information is data compensated using a Lagrange interpolation method, and the decomposed, reconstructed and data compensated electrical quantity information of each monitoring point is time-aligned to obtain standard electrical quantity information; Constructing a voltage and current video feature matrix based on the standard electrical quantity information, performing node correlation calculation on the voltage and current video feature matrix, and generating a multidimensional feature vector set representing the spatiotemporal correlation of electrical quantities; Using the node association threshold and switch state quantity in the multidimensional feature vector set, dynamically generating and updating the adjacency matrix, and reconstructing the topological graph model of the distribution network; Based on the topology model, through distributed power flow calculation and overload index analysis, the operation status mapping table and abnormal warning signal of the distribution network are output to obtain the real-time operation mode of the distribution network.

2. The method according to claim 1, characterized in that The method dynamically calculates the protection range sensitivity according to the topology structure and real-time operation mode, and uses a preset three-stage action rule to perform detection and judgment based on the protection range sensitivity and electrical quantity information to obtain a real-time protection action signal, specifically including: Based on the topology and real-time operation mode, a directional sensitivity coefficient is set for each branch of the topology, wherein the sensitivity coefficient is used to characterize the response strength of the branch to downstream faults; Constructing a sensitivity matrix based on the directional sensitivity coefficients, performing a Hadamard product operation on the sensitivity matrix and the adjacency matrix, eliminating invalid branches, and obtaining an optimized sensitivity matrix, wherein the sensitivity matrix is ​​used to determine the protection range sensitivity of each branch; Dynamically generate a three-stage protection criterion set with adaptive threshold and time delay characteristics based on the sensitivity matrix and the real-time load current in the electrical quantity information; By performing topological connectivity verification on the three-stage protection criterion set, a real-time protection action signal for the fault branch is generated.

3. The method according to claim 2, characterized in that The three-stage protection criterion set specifically includes: I set1 =K rel1 ·S·I f,max ,t set1 =0s I set3 =K rel3 ·I load ,t set3 =t set2 +β Among them, I set1 Indicates the action current threshold of stage I, I set2 Indicates the action current threshold of stage II, I set3 Indicates the action current threshold of stage III, K rel1 Represents the reliability factor used to ensure that the operating current threshold of stage I is higher than the maximum short-circuit current, K rel2 Indicates the coordination coefficient with the I-stage action current threshold, K rel3 Indicates the overload protection coefficient, S indicates the sensitivity coefficient, I f，max It represents the maximum short-circuit current of the three phases at the end of the branch, Δt represents the time step constant, α represents the sensitivity and time adjustment factor, t set1 Indicates the first stage action time, t set2 Indicates the action time of stage II, t set3 Indicates the action time of the third stage, I load Indicates the real-time load current of the branch, β indicates the time margin, I real Indicates the effective value of the current measured by the protection device in real time. Indicates the rate of change of current, t duration Indicates the duration of current exceeding the threshold; K rel1 =1.3, K rel2 =1.1,α=0.2,K rel3 =1.2,β=0.

2.

4. The method according to claim 1, wherein The method constructs an FA fault feature matrix based on action timing data, electrical quantity information, and topological structure, and uses an LSTM neural network to analyze the timing correlation of the FA fault feature matrix to generate an FA fault probability distribution map, specifically including: Align and tensor-join the action timing data, electrical quantity information, and topological structure to form the FA fault feature matrix; Construct a two-layer bidirectional LSTM network, input the FA fault feature matrix into the two-layer bidirectional LSTM network for time series correlation analysis, and obtain the fault probability value of each coordinate of the distribution network; The fault probability values ​​of each coordinate of the distribution network are output in the form of a heat map to generate a FA fault probability distribution map.

5. The method according to claim 1, wherein The method uses a fault location algorithm to perform collaborative analysis based on the real-time protection action signal and the FA fault probability distribution diagram to obtain a fault location result, specifically including: Based on the real-time protection action signal and FA fault probability distribution map, an improved matrix algorithm is used to analyze and generate a candidate set of fault areas. Based on the fault area candidate set, transient current features are extracted using wavelet packet transform to construct a decision tree classifier; Based on the fault area candidate set, a protection action logic is temporally reasoned using a Petri net to obtain a protection action logic rule base; A collaborative analysis is performed based on the decision tree classifier and the protection action logic rule base to generate a fault location result.

6. The method according to claim 1, characterized in that Based on the fault location results, a reinforcement learning algorithm is used to dynamically plan the load transfer path of the distribution network and generate the optimal isolation and restoration plan, which specifically includes: The fault location result is used as the state space, the switch operation instruction set is used as the action space, and minimizing the switch cost and load loss is used as the reward function; Based on the state space, action space and reward function, the DQN reinforcement learning algorithm is used to make action decisions and output the optimal action sequence; According to the optimal action sequence, the switch operation sequence table and the corresponding load transfer path are analyzed to generate the optimal isolation and restoration plan.

7. A rapid fault location system based on three-stage protection and FA collaborative protection, characterized in that: The system specifically includes: The data acquisition and preprocessing module is used to collect electrical quantity information of each monitoring point in the distribution network in real time, and obtain the topological structure and real-time operation mode of the distribution network by preprocessing and extracting features of the electrical quantity information; The three-stage protection module is used to dynamically calculate the protection range sensitivity based on the topology structure and real-time operation mode. Based on the protection range sensitivity and electrical quantity information, it uses the preset three-stage action rules to perform detection and judgment to obtain real-time protection action signals; The FA fault analysis module is used to collect the action time series data of each protection device in the distribution network in real time, build the FA fault feature matrix based on the action time series data, electrical quantity information and topological structure, and use the LSTM neural network to analyze the time series correlation of the FA fault feature matrix to generate the FA fault probability distribution map; The collaborative control module is used to perform collaborative analysis based on the real-time protection action signal and the FA fault probability distribution map using the fault location algorithm to obtain the fault location result; The fault self-healing module is used to dynamically plan the load transfer path of the distribution network based on the fault location results using a reinforcement learning algorithm to generate the optimal isolation and restoration plan; The method of obtaining the topology and real-time operation mode of the distribution network by preprocessing and extracting features from the electrical quantity information specifically includes: The electrical quantity information is decomposed and reconstructed using a wavelet threshold denoising algorithm, the electrical quantity information is data compensated using a Lagrange interpolation method, and the decomposed, reconstructed and data compensated electrical quantity information of each monitoring point is time-aligned to obtain standard electrical quantity information; Constructing a voltage and current video feature matrix based on the standard electrical quantity information, performing node correlation calculation on the voltage and current video feature matrix, and generating a multidimensional feature vector set representing the spatiotemporal correlation of electrical quantities; Using the node association threshold and switch state quantity in the multidimensional feature vector set, dynamically generating and updating the adjacency matrix, and reconstructing the topological graph model of the distribution network; Based on the topology model, through distributed power flow calculation and overload index analysis, the operation status mapping table and abnormal warning signal of the distribution network are output to obtain the real-time operation mode of the distribution network.

8. A computer device, characterized in that: include: A memory, a processor, and a computer program stored in the memory, which, when executed on the processor, implements the method for quickly locating faults based on three-stage protection and FA collaborative protection as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that A computer program is stored thereon, and when the computer program is executed by the processor, the method for quickly locating faults based on three-stage protection and FA coordinated protection according to any one of claims 1 to 6 is implemented.

Images