Intelligent protection and switching method and system capable of self-healing power grid fault

Abstract

The invention belongs to the technical field of power grids, and particularly relates to an intelligent protection and switching method and system capable of self-healing power grid faults, and the method comprises the steps: integrating a verified actual connection graph, a power supply direction, line impedance and capacity information into a weighted directed graph through building a dynamic topology database containing electrical parameters, and carrying out the self-healing of the power grid faults; therefore, more accurate fault positioning is supported. Particularly, when voltage drop and current sudden increase are detected, the dynamic topology database can be rapidly utilized to analyze the change rule of the power flow direction before and after the fault, the power interruption point is accurately judged, and the fault propagation path is confirmed. According to the fault propagation path, the upstream and downstream operable switches are selected to execute opening operation, and rapid isolation of the fault section is realized. According to the method, the fault positioning and isolating time is greatly shortened, the power failure range is reduced, and the self-healing capability and the power supply reliability of the power grid are improved.

Description

Technical Field

[0001] This invention belongs to the field of power grid technology, specifically relating to an intelligent protection and switching method and system capable of self-healing from power grid faults. Background Technology

[0002] In existing power grid protection and switching systems, traditional methods based on relay protection are commonly used for fault detection and isolation. This method primarily relies on pre-set current or voltage thresholds to trigger circuit breaker operation, thereby isolating the faulty section. However, traditional solutions have several significant shortcomings: First, they often only identify whether a fault has occurred, but cannot accurately determine the specific location of the fault; second, due to the lack of real-time dynamic topology information, these systems struggle to effectively cope with complex network structures and variable operating modes; finally, when a fault occurs, traditional methods may take a considerable amount of time to locate and isolate the faulty section, which not only affects power supply reliability but may also lead to wider power outages.

[0003] Therefore, how to quickly and accurately locate the precise location of a power grid fault and promptly and effectively isolate the faulty section in order to reduce the scope and duration of power outages and improve the self-healing capability and power supply reliability of the power grid is an urgent problem to be solved. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent protection and switching method and system capable of self-healing power grid faults. This method not only greatly shortens the time for fault location and isolation and reduces the scope of power outages, but also improves the self-healing capability and power supply reliability of the power grid, thereby solving the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: an intelligent protection and switching method capable of self-healing from power grid faults, comprising: The voltage, current and power data of each power distribution node are collected synchronously, and the current actual connection diagram is generated by combining the switch status information. The voltage, current and power data are used to verify the energization status of the line segments in the actual connection diagram. Identify the power supply direction under the current operating mode, and determine the power input source of each node based on the verified actual connection diagram; Establish a dynamic topology database containing electrical parameters, and integrate the verified actual connection diagram, power supply direction, line impedance, and capacity information into a weighted directed graph, where nodes represent power distribution equipment and edges represent line connection relationships. The spatiotemporal distribution characteristics of voltage dips and current surges are detected, and the feeder section corresponding to the location of the voltage dip is located using a dynamic topology database; Analyze the changes in power flow direction before and after the fault, and use the upstream and downstream relationships of the feeder section in the dynamic topology database to determine the power interruption point; The continuity of fault paths is verified by combining a dynamic topology database, and the fault propagation path is confirmed by the topological connectivity between the power interruption point and the actual connection graph. Identify the faulty section and isolate the affected area. Based on the fault propagation path, select upstream and downstream operable switches to perform tripping operations.

[0006] Preferably, the synchronous acquisition of voltage, current, and power data at each power distribution node further includes: Three-phase electrical quantities are acquired at a millisecond-level frequency using a measurement unit with time synchronization function; The collected data is then linked to a timestamp and uploaded to the regional control center. The voltage drop threshold is set at 70% of the rated voltage, and the current rise rate threshold is five times the rated current per millisecond. Calculate the direction of active power for each line segment and record the changes before and after the fault.

[0007] Preferably, the step of generating the current actual connection diagram by combining the switch status information further includes: By overlaying the physical topology with the real-time switch remote signaling status, an actual connectivity diagram reflecting the current power grid connection status is formed. Based on the actual connected graph, traverse all edges to analyze the power flow direction and determine the power convergence point and flow direction. After completing the power flow analysis, nodes with only one power inflow direction are marked as upstream power nodes; Nodes with multiple power input directions are identified as connection points or ring points to distinguish complex network structures.

[0008] Preferably, identifying the power supply direction under the current operating mode further includes: By traversing all edges in the actual connected graph and analyzing the power flow towards the convergence point, the single power input source of each node can be determined. If a node has multiple power input directions, it is marked as a tie point or a ring point.

[0009] Preferably, establishing a dynamic topology database containing electrical parameters further includes: Write the impedance parameters, rated capacity, and current load rate information of each line into the attribute fields of the corresponding edge in the weighted directed graph; The load type and importance level of each node are written into the attribute field of the corresponding vertex to form a multi-attribute topology model that supports path search and security verification.

[0010] Preferably, the detection of the spatiotemporal distribution characteristics of voltage dips and current surges further includes: Set voltage sag threshold and current rise rate threshold; When the voltage of a node falls below a threshold and is accompanied by a sudden increase in current, the node is determined to be in the fault-affected area. The time and spatial location of such events are recorded to form a preliminary set of suspected fault points.

[0011] Preferably, the analysis of the power flow direction change pattern before and after the fault further includes: Line segments where the power direction was positive before the fault but returned to zero or reversed after the fault were marked as suspected fault paths; Find the shortest connected path between the starting point of the voltage drop and the point of maximum current surge on the suspected fault path. The shortest connected path is used as a candidate fault propagation path for subsequent verification.

[0012] Preferably, determining the faulty section and isolating the scope of the fault's impact further includes: Locate the power-side switch and load-side switch closest to the fault point along the fault propagation path; Send a trip command to disconnect the two power supply side switches and the load side switch; The fault is confined to the line segment between the two power-side switches and the load-side switch; The final confirmed fault section is used for subsequent power supply path reconfiguration.

[0013] Preferably, the step of verifying the continuity of fault paths using a dynamic topology database further includes: Locate the two nearest operable switches along the fault path: one near the power supply side and the other near the load side. Issue trip commands to both switches to disconnect them, thereby confining the fault to the line segment between them. This line segment is the finally determined fault section.

[0014] On the other hand, this invention proposes an intelligent protection and switching system capable of self-healing from power grid faults, comprising: A measurement unit with time synchronization function is used to synchronously collect voltage, current and power data of each power distribution node; The regional control center is used to generate the current actual connection diagram by combining switch status information, and to verify the energization status of the line segments in the actual connection diagram by referencing voltage, current and power data. The power direction identification unit is used to identify the power supply direction under the current operating mode and determine the power input source of each node based on the verified actual connection diagram. A dynamic topology database is used to store power grid connection information containing electrical parameters and integrate verified actual connection diagrams, power supply direction, line impedance, and capacity information into a weighted directed graph. The fault feature detection unit is used to detect the spatiotemporal distribution characteristics of voltage dips and current surges, and to locate the feeder section corresponding to the location where the voltage dip occurs using a dynamic topology database. The power flow direction analysis unit is used to analyze the change pattern of power flow direction before and after the fault, and to determine the power interruption point by referring to the upstream and downstream relationship of the feeder section in the dynamic topology database. The fault path verification unit is used to verify the continuity of fault paths by combining a dynamic topology database and to confirm the fault propagation path by the topological connectivity between the power interruption point and the actual connection graph. The fault isolation execution unit is used to determine the faulty section and isolate the scope of the fault's impact, and select upstream and downstream operable switches to perform tripping operations based on the fault propagation path.

[0015] Technical effects and advantages of the present invention: The intelligent protection and switching method and system capable of self-healing power grid faults proposed in this invention have the following advantages compared with the prior art: This invention generates a current actual connection diagram by synchronously collecting voltage, current, and power data from each distribution node and combining this data with switch status information. This data is then used to verify the energization status of line segments, ensuring the real-time nature and accuracy of the power grid topology information. Furthermore, this method establishes a dynamic topology database containing electrical parameters, integrating the verified actual connection diagram, power supply direction, line impedance, and capacity information into a weighted directed graph, thereby supporting more precise fault location. Especially when voltage dips and current surges are detected, the dynamic topology database can be quickly used to analyze the changes in power flow before and after the fault, accurately determining the power interruption point and confirming the fault propagation path. Finally, based on the fault propagation path, upstream and downstream operable switches are selected to perform tripping operations, achieving rapid isolation of the faulty section. This method not only significantly shortens the time for fault location and isolation and reduces the power outage area, but also improves the self-healing capability of the power grid and the reliability of power supply. Attached Figure Description

[0016] Figure 1 The flowchart shows the intelligent protection and switching method for self-healing of power grid faults according to the present invention. Figure 2 This is a block diagram of the intelligent protection and switching system of the present invention, which is capable of self-healing from power grid faults. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] This invention provides, for example Figure 1 This paper presents an intelligent protection and switching method for self-healing power grid faults. The aim is to achieve intelligent reconfiguration and seamless switching of power supply paths in non-faulty areas after a power grid fault occurs, ensuring continuous power supply to critical loads. This method achieves rapid response and power supply continuity assurance of the distribution network under fault disturbances through distributed real-time state perception, accurate fault area identification, dynamic evaluation of power supply paths, collaborative generation of switching commands, closed-loop monitoring of the execution process, and post-recovery state optimization. The entire process relies on dynamic modeling of the power grid topology and collaborative analysis of electrical quantity information. Combined with multi-dimensional data such as voltage, current, and power flow, it constructs a decision-making system for self-healing control, ensuring that critical users can maintain stable power supply during fault periods. Specifically, it includes: By synchronously collecting voltage, current, and power data from each distribution node and combining this data with switch status information to generate the current actual connection diagram, and then using the voltage, current, and power data to verify the energization status of line segments in the actual connection diagram, the accuracy of assessing the grid's operational status can be effectively improved, allowing for the timely detection of potential problems or anomalies. This real-time monitoring and verification mechanism reduces the risk of misjudgments due to topology errors, improves the reliability and response speed of fault detection, and helps to quickly locate problem areas and take corresponding measures, thereby enhancing the stability and security of the entire power system. Furthermore, this method supports more efficient resource management and scheduling optimization, providing a solid foundation for the dynamic operation of the smart grid.

[0019] The process includes: after synchronously collecting voltage, current, and power data from each distribution node, the following steps are also taken: using measurement units with time synchronization capabilities to acquire three-phase electrical quantities at millisecond-level frequencies; binding the collected data with timestamps and uploading it to the regional control center; setting a voltage drop threshold of 70% of the rated voltage and a current rise rate threshold of five times the rated current per millisecond; calculating the direction of active power for each line segment and recording changes before and after a fault. Setting voltage drop and current rise rate thresholds and monitoring accordingly allows for the rapid identification of abnormal conditions in the power grid, such as short circuits or overloads, enabling timely measures to prevent greater losses. Calculating the direction of active power and recording changes before and after a fault not only helps in accurately locating the fault point but also provides data analysis support for the stable operation of the power system, improving the system's self-protection capabilities and recovery speed.

[0020] After generating the current actual connection diagram by combining switch status information, the process also includes: overlaying the physical topology with the real-time switch remote signaling status to form an actual connectivity diagram that reflects the current grid connection status; based on the actual connectivity diagram, traversing all edges to analyze power flow direction and determine power convergence points and flow directions; after completing the power flow direction analysis, marking nodes with only one power inflow direction as upstream power source nodes; and identifying nodes with multiple power input directions as tie points or ring network points to distinguish complex network structures.

[0021] It can accurately construct a connectivity structure reflecting the real-time operating status of the power grid, effectively distinguishing normal power supply paths from potential ring networks or backup connections. Power flow analysis based on the actual connectivity diagram can accurately identify power sources and load receiving paths, avoiding fault location errors caused by topology misjudgments. Clearly marking upstream power nodes helps quickly pinpoint the power supply origin, providing direction for fault isolation and recovery strategies. Identifying tie points and ring network points enhances adaptability to complex distribution network operation modes, preventing unintended parallel or islanded operation caused by misoperation, and improving the system's control safety and self-healing reliability in multi-power source and multi-tie scenarios.

[0022] Identify the power supply direction under the current operating mode, and determine the power input source of each node based on the verified actual connection diagram; After identifying the power supply direction under the current operating mode, the process also includes: determining the single power input source of each node by traversing all edges in the actual connected graph and analyzing the power flow convergence point; if a node has multiple power input directions, it is marked as a connection point or a ring network point.

[0023] This technology enables precise identification and labeling of the power supply sources for each node in the power grid, enhancing the understanding and control of the grid's operational status. Clearly defining the power input source of each node helps optimize grid dispatching decisions and improve fault handling efficiency, particularly in quickly locating fault sources and determining isolation ranges. Identifying nodes with multiple power input directions as tie points or ring points effectively supports load balancing, fault transfer, and optimized operating strategies in complex network structures, avoiding safety risks and system instability caused by operational errors.

[0024] Establish a dynamic topology database containing electrical parameters, and integrate the verified actual connection diagram, power supply direction, line impedance, and capacity information into a weighted directed graph, where nodes represent power distribution equipment and edges represent line connection relationships. The process includes establishing a dynamic topology database containing electrical parameters, and further involves: writing the impedance parameters, rated capacity, and current load rate information of each line into the attribute fields of the corresponding edges in a weighted directed graph; and writing the load type and importance level of each node into the attribute fields of the corresponding vertices, forming a multi-attribute topology model that supports path search and security verification. This model can adapt to complex distribution network structures and provides reliable support for switching operations under both normal operation and fault conditions, ensuring the stability and flexibility of the power grid under various operating modes.

[0025] The system detects the spatiotemporal distribution characteristics of voltage dips and current surges, and uses a dynamic topology database to locate the feeder section corresponding to the location of the voltage dip; this significantly improves the speed and accuracy of fault detection.

[0026] The process includes detecting the spatiotemporal distribution characteristics of voltage dips and current surges, as well as setting voltage dip thresholds and current rise rate thresholds. When the voltage of a node falls below the threshold and is accompanied by a sudden current surge, the node is determined to be in the fault-affected area. The time and spatial location of such events are recorded to form a preliminary set of suspected fault points. Setting voltage dip thresholds and current rise rate thresholds enables rapid identification of abnormal situations in the power grid, ensuring a response in the early stages of a fault.

[0027] Analyze the changes in power flow direction before and after the fault, and use the upstream and downstream relationships of the feeder section in the dynamic topology database to determine the power interruption point; The process includes, after analyzing the changes in power flow before and after the fault, marking line segments where the power direction was positive before the fault and zero or reversed after the fault as suspected fault paths; finding the shortest connected path between the voltage drop initiation point and the current surge point on the suspected fault path; and using the shortest connected path as a candidate fault propagation path for subsequent verification.

[0028] By analyzing the changes in power flow before and after a fault, and combining this with the upstream and downstream relationships in the dynamic topology to identify power interruption points, we can accurately pinpoint the candidate range of fault propagation paths, effectively eliminate interference from non-faulty areas, improve the accuracy and robustness of fault segment judgment, and provide a reliable basis for subsequent rapid isolation and power restoration.

[0029] The continuity of fault paths is verified by combining a dynamic topology database, and the fault propagation path is confirmed by the topological connectivity between the power interruption point and the actual connection graph. After verifying the continuity of the fault path using a dynamic topology database, the process also includes: finding the two nearest operable switches on the fault path: one near the power supply side and the other near the load side, and issuing tripping commands to these two switches to disconnect them, thereby confining the fault to the line segment between them, which is the finally determined fault section.

[0030] By verifying the continuity of the fault path and combining it with topological connectivity to accurately locate the faulty section, the accuracy of isolation operations can be ensured and the power outage area minimized, avoiding accidental disconnection of non-faulty lines and effectively improving the reliability of fault handling and the self-healing efficiency of the power grid.

[0031] Identify the faulty section and isolate the affected area; select upstream and downstream operable switches to perform tripping operations based on the fault propagation path. After identifying the faulty section and isolating the scope of the fault's impact, the process also includes: locating the power supply side switch and load side switch closest to the fault point along the fault propagation path; issuing tripping commands to the two power supply side switches and load side switches to disconnect them; confining the fault to the line segment between the two power supply side switches and the load side switch; and forming the finally confirmed faulty section for subsequent power supply path reconstruction.

[0032] By locating the faulty section and coordinating the tripping of upstream and downstream switches, the faulty area can be quickly and reliably isolated, effectively blocking the path of fault propagation, maximizing the protection of continuous power supply to non-faulty areas, and providing clear boundary conditions and a safety basis for subsequent restoration operations.

[0033] On the other hand, this invention proposes an intelligent protection and switching system capable of self-healing from power grid faults, such as... Figure 2 As shown, it includes: A measurement unit with time synchronization function is used to synchronously collect voltage, current and power data of each power distribution node; The regional control center is used to generate the current actual connection diagram by combining switch status information, and to verify the energization status of the line segments in the actual connection diagram by referencing voltage, current and power data. The power direction identification unit is used to identify the power supply direction under the current operating mode and determine the power input source of each node based on the verified actual connection diagram. A dynamic topology database is used to store power grid connection information containing electrical parameters and integrate verified actual connection diagrams, power supply direction, line impedance, and capacity information into a weighted directed graph. The fault feature detection unit is used to detect the spatiotemporal distribution characteristics of voltage dips and current surges, and to locate the feeder section corresponding to the location where the voltage dip occurs using a dynamic topology database. The power flow direction analysis unit is used to analyze the change pattern of power flow direction before and after the fault, and to determine the power interruption point by referring to the upstream and downstream relationship of the feeder section in the dynamic topology database. The fault path verification unit is used to verify the continuity of fault paths by combining a dynamic topology database and to confirm the fault propagation path by the topological connectivity between the power interruption point and the actual connection graph. The fault isolation execution unit is used to determine the faulty section and isolate the scope of the fault's impact. Based on the fault propagation path, it selects upstream and downstream operable switches to perform tripping operations. In addition, when the aforementioned system's constituent units are executed, they are also used to implement other steps of the aforementioned intelligent protection and switching method capable of self-healing from power grid faults, as follows: Step 1: Real-time acquisition of power grid operating status and establishment of dynamic topology model During normal operation of the power grid or at the initial stage of a disturbance, it is necessary to first synchronously collect electrical quantity information from all key nodes in the entire network, and then construct a dynamic topology model reflecting the current actual connection relationships based on the collected data. This model not only includes physical connection relationships, but also integrates switch status, equipment operation status, and load distribution characteristics, providing basic support for subsequent fault identification and path reconfiguration.

[0034] Step 1.1: Synchronously collect voltage, current, and power data from each power distribution node: Measurement units with time synchronization capabilities are deployed at the feeder ends, tie switches, sectionalizing switches, and key load connection points of the distribution network to acquire three-phase voltage at millisecond-level sampling frequencies. Three-phase current and instantaneous active and reactive power This data is uploaded to the regional control center via a high-speed communication network. The acquisition process ensures consistent timestamps to facilitate subsequent cross-node phase comparisons and fault direction determination. This forms a "perception layer" covering the entire network, providing the initial basis for subsequent analysis. Without this data foundation, all subsequent judgments would lack objective support.

[0035] Step 1.2: Generate the current actual connection diagram based on the switch status information: Using the time synchronization data obtained in step 1.1, the remote signaling status of each automated switch (such as pole-mounted circuit breakers and ring main unit load switches) is also retrieved. ,in =1 indicates a closed loop. =0 indicates disconnection. The physical topology of the distribution network (i.e., the line connection diagram) is overlaid with the real-time switch status to generate a connectivity graph G=(V,E) of the current actual energized path, where vertex set V represents busbars or load nodes, and edge set E represents line segments in the conducting state. This eliminates the discrepancy between the "nominal topology" and the "actual topology," avoiding misjudgments due to false alarms from switches or remote control failures. For example, if a sectionalizing switch is closed on the drawing but actually tripped due to a fault, its disconnected state must be reflected in the model.

[0036] Step 1.3: Identify the power supply direction under the current operating mode: Referencing the connected graph generated in step 1.2 Based on the power flow data collected in step 1.1, the active power direction is calculated for each line segment. ,in This represents the active power flowing from node i to node j. By traversing all edges in the graph and analyzing the power flow convergence point, we determine which substation or distributed power source supplies each feeder segment. If a node has a unique power inflow direction, it is marked as an "upstream power source node"; if multiple directions of input exist, it is identified as a "connection point" or "ring point." Clarifying the power supply distribution of the power grid provides a directional basis for subsequent assessment of the fault impact range. For example, when a section of the line loses power, this step can be used to determine its original power source, and then determine whether it can be restored through other paths.

[0037] Step 1.4: Establish a dynamic topology database containing electrical parameters: Referencing the connected graph from step 1.2 In conjunction with the power supply direction information from step 1.3, the impedance parameters of each line are further determined. Rated capacity Current load rate Information such as load type (e.g., industrial, commercial, residential) and importance level of each node are appended to the edge attributes of the graph. (1 being the highest) Write the vertex attributes. This ultimately forms a weighted directed graph. =(V,E,W), where the weight set W contains multi-dimensional information such as impedance, capacity, load factor, and importance. This provides complete input conditions for subsequent path evaluation. For example, when selecting a reconfiguration path, heavily loaded lines can be avoided or low-reliability sections can be bypassed, thereby improving operational safety after switching.

[0038] Step 2: Identify faulty sections based on electrical quantity mutations and topology consistency verification. After completing the dynamic topology modeling, when the power grid experiences abnormal operating conditions, it is necessary to quickly identify the specific section where the fault occurred. This step analyzes the abrupt changes in voltage and current, and performs cross-validation by combining the consistency of the power supply path in the topology model to eliminate false alarm interference and achieve accurate location of the fault section.

[0039] Step 2.1: Detect the spatiotemporal distribution characteristics of voltage dips and current surges: Using the synchronous electrical quantity data collected in step 1.1 of step one, a voltage drop threshold is set. Current rise rate threshold When the voltage at a certain node is lower than And accompanied by a sudden increase in current exceeding When this occurs, the node is determined to be within the fault-affected area. The time of occurrence of such events is recorded. Spatial location This forms a preliminary set of suspected fault points. The typical electrical characteristics of short-circuit faults can be used to quickly pinpoint suspicious areas. For example, a metallic short circuit causes a sharp drop in voltage and a surge in current; this combination of characteristics is highly distinctive and can be differentiated from sudden load changes or lightning strikes.

[0040] Step 2.2: Analyze the changes in power flow direction before and after the fault: Refer to the set of suspected fault points identified in step 2.1 This involves retrieving power direction data from relevant lines before and after the fault occurred. Under normal circumstances, power should flow in a single direction from the power source to the load; however, during a short circuit, the power downstream of the fault point may reverse or return to zero, while the power upstream of the fault line increases significantly. Let's assume a certain line segment... The power direction before the fault was After the fault If satisfied If the fault is detected, the line is determined to be on a fault path. Identifying the fault propagation path through interruptions or reversals in power flow enhances the physical plausibility of the judgment. For example, if a line previously had positive power output but lost power after a fault, it indicates that its downstream areas have lost power, which may be the location of the fault.

[0041] Step 2.3: Verify the continuity of the fault path using a dynamic topology model: Referencing the abnormal power flow line set identified in step 2.2 This is mapped to the dynamic topology graph established in step 1.4 of step one. Starting from the power node in the diagram, trace the path along the normal power supply direction to check if there is a path that originates from the power source and passes through... All abnormal routes and final destinations Analyze the continuous paths of suspected fault points. If a path exists and is unique, it is confirmed as the actual fault propagation path; if multiple paths meet the criteria, further analysis is required. Utilize topological logic to eliminate isolated anomalies, preventing misjudgments due to local measurement errors. For example, if the voltage at a node drops but the power of its upstream line remains unchanged, it may indicate a PT disconnection rather than a genuine short circuit.

[0042] Step 2.4: Identify the faulty section and isolate the affected area: Referring to the fault propagation path identified in step 2.3, locate the two nearest operable switches along this path: one located upstream of the fault point (close to the power supply side), denoted as... The other one is located downstream (closer to the load side), denoted as... Sending a trip command to both switches disconnects them, thus limiting the fault to [the area where the fault is located]. and Within the line segment between [the specified points]. This line segment is the finally determined faulty section. This achieves physical isolation of the fault, preventing the spread of fault current and creating conditions for power restoration in non-faulty areas. For example, after disconnection, lines outside the faulty section are no longer affected by the short circuit, and voltage can gradually recover, providing a stable background for subsequent switching.

[0043] Step 3: Evaluate the set of reconfigurable power supply paths for non-faulty regions After the faulty section is successfully isolated, the power supply path for the remaining non-faulty areas needs to be replanned. This step, based on a dynamic topology model, searches for all possible alternative power supply paths and performs a comprehensive evaluation of each path's safety, capacity margin, and load priority to form a set of selectable paths.

[0044] Step 3.1: Identify the power outage area and its associated load nodes: Refer to the fault section identified in step 2.4 of step two. In dynamic topology graph In China, with Starting from the original power supply direction, perform a breadth-first search (BFS) downstream to find the set of all nodes that have lost power connection due to fault isolation. These nodes are the "power-outage load areas" that require power restoration. Simultaneously, the load importance level of each node is extracted from the vertex attributes in step 1.4. Number of Level 1 loads within the statistical area With total load capacity Clearly define the scope and priority of restoration targets. For example, if a power outage area includes a hospital or data center, its restoration urgency is higher than that of ordinary residential areas.

[0045] Step 3.2: Search for available backup power sources and their connection paths: Refer to the power loss area identified in step 3.1 In the topology graph The system identifies all unaffected power sources capable of supplying power, including backup feeders at the main substation, outlets connected to adjacent feeders via tie switches, and distributed power sources (such as photovoltaic + energy storage systems) with islanding capabilities. For each available power source... Using Dijkstra's shortest path algorithm to search from arrive Find the path with the minimum number of hops for any node, and check if all switches on the path are in a remotely controllable state. Record all feasible paths to form a candidate set. Expanding power sources breaks the dependence on a single power source. For example, drawing power from a nearby feeder via a tie switch enables cross-regional support.

[0046] Step 3.3: Calculate the load-carrying capacity and voltage stability of each candidate path: Referencing the candidate path set generated in step 3.2 For each path Calculate its overall impedance And according to the rated capacity of each line on the route. Determine bottleneck capacity At the same time, estimate the total load that this path needs to bear. Calculate the load rate Furthermore, using the simplified voltage drop formula: ; in For active and reactive power transmitted along the path, The total resistance and reactance of the path are given. This is the rated voltage. If the path does not meet the safety operation requirements, it is determined that the path does not meet the requirements. Paths with overload or excessive voltage drop are eliminated to ensure power quality after power is restored. For example, long-distance low-cross-section lines under heavy load may cause the terminal voltage to be too low, affecting equipment operation.

[0047] Step 3.4: Generate a set of feasible power supply paths that satisfy safety constraints: Referring to the evaluation results of step 3.3, all paths that pass the load-bearing capacity and voltage stability verification are retained to form a set of feasible paths. .for For each path, add its corresponding power type and switching delay estimate. (Positively correlated with path length), number of critical loads supported These attributes provide clear input options for the next step of route optimization. For example, if a route has sufficient capacity but a long delay, it can be used as an alternative in an emergency.

[0048] Step 4: Generate the optimal switching instruction sequence and coordinate its execution. After obtaining a set of feasible power supply paths, the optimal path needs to be selected from them, and a series of switching operation instructions need to be generated to ensure a smooth and shock-free switching process, and to prioritize the rapid recovery of high-priority loads.

[0049] Step 4.1: Sort feasible paths according to load priority and path performance: Referencing the feasible path set generated in step 3.4 of step three. For each path Calculate the overall score: ; in Weighting coefficients (satisfying) These respectively reflect the importance attached to critical load support, load margin, and switching speed; The maximum value in the set. (By...) sorted in descending order To obtain the optimal path To enable multi-objective trade-off decision-making. For example, in disaster emergency response modes, it can improve... Prioritize the restoration of critical facilities.

[0050] Step 4.2: Define the sequence of switching operations from the current state to the target path: Refer to the optimal path selected in step 4.1 The difference between the current switch state and the expected state under the target power supply mode is analyzed. To avoid large currents generated when the ring network closes, the "disconnect before closing" principle is adopted: first, disconnect any residual connections on the original power supply path that no longer have current (such as tie switches), and then close the loop. Interconnectors or sectionalizing switches along the path. The operation sequence proceeds from the side furthest from the power source to the side closest to the power source, forming a command sequence. Each operation includes the switch number, action type (close / open), and execution sequence. This ensures electrical safety during the switching process. For example, closing before disconnecting could cause two power supplies to operate in parallel, triggering protection mechanisms.

[0051] Step 4.3: Predict voltage fluctuations and inrush currents during the switching process: Using the operation sequence generated in step 4.2, simulate the impact of each closing operation on the system. Considering the line-to-ground capacitance, inrush current will be generated at the moment of closing. Its amplitude can be estimated as follows: ; in For line wave impedance, This represents the voltage phase at the moment of closing the circuit breaker. This is the line impedance angle. If... If the downstream protection setting is exceeded, the closing phase angle can be adjusted or a current-limiting resistor can be inserted. This reduces the stress impact on equipment during switching. For example, closing near the voltage zero-crossing point can suppress inrush current peaks.

[0052] Step 4.4: Issue cooperative control instructions with timing constraints: Referring to the impact assessment results in step 4.3, if a high inrush risk exists, a delay or synchronization condition should be added to the command; otherwise, it should be packaged into a control message according to the original sequence and sent to each field terminal through a secure communication channel. The command includes an execution time window, confirmation feedback requirements, and an anomaly rollback plan. All operations must be performed within... The process is completed internally, enabling near-seamless power transfer. It ensures coordination and real-time performance across multiple points of operation. For example, if a switch fails to confirm its action, the system will pause subsequent commands and activate a backup plan.

[0053] Step 5: Monitor the switchover process and implement closed-loop feedback adjustment. After the power supply path switching command is issued, the system enters the dynamic execution phase. The core task of this phase is to ensure that each operation is completed as expected and to verify the actual operating status in real time. If any deviation or anomaly is detected, corrective measures must be taken immediately to prevent overall recovery failure due to local failure. This step constructs a complete execution monitoring closed loop through multi-level status feedback, electrical quantity comparison, anomaly identification, and dynamic intervention mechanisms to ensure the reliability and safety of the switching process.

[0054] Step 5.1: Receive real-time confirmation of switch action and electrical quantity feedback: Using the coordinated control command sequence generated in step 4.4 of step four, the system continuously monitors feedback information from each field terminal within the command execution window. After each automatic switch completes its opening or closing action, it transmits its final position signal (remote signaling) and operation completion timestamp back via the communication network. Simultaneously, synchronous measurement devices deployed at key nodes upload three-phase voltage data at a frequency of 1000 points per second. Data. This data is centrally stored in the time-series database of the regional control center and aligned with the expected action times in the command sequence. An "command-execution-feedback" information loop is established to ensure the traceability of control actions. For example, if a handrail switch is supposed to close at t = 100ms, but fails to report a closed signal by t = 300ms, it can be preliminarily determined that the device has a communication delay or mechanical fault.

[0055] Step 5.2: Verify the voltage recovery status and load connection status of the power outage area: Using the electrical quantity data received in step 5.1, the focus is on analyzing the original power loss area. Voltage recovery trajectories for each load node within the system. (Selection) A representative set of nodes in Calculate its average voltage amplitude Set recovery criteria: If the last switch operation is completed... Inside, If the voltage recovery lasts for more than 100ms, it is considered that the voltage has been effectively restored. Further check whether the current at each node has risen from zero to the normal load level, and whether the power factor is within a reasonable range (0.85–1.0), to confirm that the load is actually energized and not just charging under no-load conditions. Distinguish between "formal recovery" and "substantive recovery." For example, although a line is energized, the load may still be unable to use power because the downstream circuit breaker has not automatically reclosed. In this case, a supplementary closing command needs to be triggered.

[0056] Step 5.3: Check for unexpected circulating currents or overload phenomena: Using the voltage recovery data confirmed in step 5.2, a power flow scan is further performed on all network segments. For each line (i,j), its current transmission power is calculated. and the long-term allowable current carrying capacity of the line. corresponding capacity Comparison. If If the warning persists for more than 30 seconds, it will be marked as an "overload warning".

[0057] More importantly, circulating current detection is crucial: select all line sections where tie switches are located, and if the power supplies at both ends come from different substations, and active power is measured... Continuing flow even when there is no load demand If a loop current is detected, it is determined that a loop current exists. At this point, using the topology model established in step 1.4, the power source is traced in reverse to determine if parallel operation is caused by the simultaneous closure of multiple tie points. This helps prevent secondary risks. For example, if two 10kV feeders are connected simultaneously through two or more tie switches, a low-impedance loop may be formed, leading to a doubling of short-circuit current and endangering equipment safety.

[0058] Step 5.4: Dynamically adjust operating parameters based on actual operating status: Referring to the overload or circulating current issues identified in step 5.3, a local regulation mechanism is initiated. For heavily loaded lines, priority is given to utilizing distributed power sources along the line (such as rooftop photovoltaics and energy storage systems) to increase local output and reduce the burden on the main power supply. Let the current load of a certain line be... Its bottleneck capacity is The power that needs to be compensated is: ; If there are schedulable distributed units along the route Its remaining available capacity is Then, they will be deployed according to priority. until For circulating current issues, the connecting switch furthest from the load center in the loop is selected, and a tripping command is issued to break the loop structure. All adjustment actions are completed within 1 second, and the effect is verified again through steps 5.1 to 5.3, forming a closed loop of "monitoring-judgment-adjustment-remonitoring". This achieves adaptive optimization of the operating state. For example, by supporting the local power supply, voltage collapse caused by a single path carrying too much load can be avoided.

[0059] Step Six: Complete state optimization and system normalization after power restoration. After the switching process is completed, although the system has restored power supply, it may be operating in a temporary emergency state. To improve long-term operating efficiency, reduce equipment wear and tear, and prepare for the next disturbance, the current state needs to be systematically organized and normalized. This step, through four stages—operating state solidification, resource release, file updating, and system self-checking—smoothly transitions the power grid from "emergency mode" to "normal mode," ensuring the sustainability of overall operation.

[0060] Step 6.1: Confirm that all loads are running stably and record the final topology status: Using the operational data adjusted in step 5.4 of step five, continuously monitor key indicators across the entire network at least The monitoring content includes: the voltage fluctuation range of each bus (which should be controlled within...). Internal), system frequency deviation (should be less than) The main power supply line load rate (should be below 80%), continuous power supply time for important loads, etc., are considered. If all indicators meet the stability conditions, the current switch state combination, power supply path, distributed power output level, and other information are packaged into an operational status snapshot. This snapshot contains a complete dataset including topology diagrams, voltage amplitudes at each node, power flow distribution on major lines, and protection setting configurations. It establishes a new, long-term sustainable operating baseline. For example, if a feeder is routinely powered by an adjacent line via a tie switch, this should be considered the new primary power supply method, and protection coordination should be adjusted accordingly.

[0061] Step 6.2: Release temporarily enabled backup resources and communication channels: Referencing the runtime snapshot generated in step 6.1 The system identifies which resources were temporarily used during the recovery process. For example, a diesel generator provided emergency support before the main power supply was restored, or a dedicated wireless communication channel was used to ensure priority transmission of switching commands. The system queries the load margin of the current main power supply path; if it meets the requirements... If no other disturbances are observed, a withdrawal command is issued to the temporary resources. For power generation resources, their output power is gradually reduced until they are disconnected from the grid; for communication resources, their priority is restored to normal levels. The release process is gradual to avoid fluctuations caused by sudden changes. Unnecessary operating costs are reduced. For example, prolonged operation of diesel engines not only consumes fuel but also accumulates mechanical fatigue; timely shutdown can extend their service life and ensure availability for the next emergency.

[0062] Step 6.3: Update the power grid operation records and historical event logs: Referencing the runtime snapshot of step 6.1 A complete event archive file is generated, along with the operation log of the entire self-healing process. This file contains the following: fault occurrence time and location, preliminary fault type assessment (e.g., phase-to-phase short circuit, single-phase grounding), isolation operation sequence, basis for power restoration path selection, total switching time, list of affected loads and their outage duration, list of distributed power sources participating in regulation, and details of the final operating mode change. This file is written to the power grid historical database via a security interface, triggering the topology management module to update the current operating mode diagram. Simultaneously, the system automatically compares this event with similar historical cases, extracting common features for subsequent rule optimization, thus forming a knowledge accumulation mechanism. For example, if a region experiences repeated recovery delays due to the slow operation of the same tie switch, redundant control channels can be configured for it in subsequent contingency plans.

[0063] Step 6.4: Perform system self-check and prepare for the next round of monitoring: Using the updated runtime archive from step 6.3, initiate a system-wide health assessment. Perform usage intensity statistics for all equipment participating in this self-healing process, including cumulative switch operations, protection device activations, and communication terminal data throughput. If a piece of equipment's operation count approaches 80% of its design life, mark it as a "key concern" in the maintenance management system and recommend scheduling an inspection during the next power outage window. Simultaneously, switch the monitoring mode of the regional control center from "Event Response" to "Normal Scan," and reload the latest runtime diagram. As a baseline model, the real-time data acquisition and topology update tasks from step one continue to be executed. This enables the system to maintain itself and remain continuously available. For example, through periodic self-checks, potential fault points can be identified in advance, preventing self-healing failure due to equipment malfunction during subsequent disturbances.

[0064] In summary, this embodiment, from initial state perception to final system normalization, uses the output of the previous stage as input for each step, forming an inseparable decision-making and execution chain. By introducing multi-dimensional real-time monitoring, anomaly identification, dynamic adjustment, resource management, and state archiving mechanisms, the robustness and sustainability of the system under complex operating conditions are significantly improved.

[0065] The entire solution not only focuses on the outcome of "restoring power supply," but also emphasizes the safety, economy, and maintainability of the restoration process. Through the fusion analysis of multi-source information such as voltage, current, power, and switch status, combined with the physical characteristics of the power grid and operational experience, intelligent reconfiguration and seamless switching of power supply paths in non-faulty areas are achieved. Critical loads can regain stable power within hundreds of milliseconds after a fault occurs, minimizing the impact of power outages.

[0066] In addition, the system has good scalability and can adapt to modern power distribution network environments with a high proportion of distributed power sources, interconnection of multiple voltage levels, and complex ring network structures.

[0067] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart protection and switching method capable of self-healing from power grid faults, characterized in that, include: The voltage, current and power data of each power distribution node are collected synchronously, and the current actual connection diagram is generated by combining the switch status information. The voltage, current and power data are used to verify the energization status of the line segments in the actual connection diagram. Identify the power supply direction under the current operating mode, and determine the power input source of each node based on the verified actual connection diagram; Establish a dynamic topology database containing electrical parameters, and integrate the verified actual connection diagram, power supply direction, line impedance, and capacity information into a weighted directed graph, where nodes represent power distribution equipment and edges represent line connection relationships. The spatiotemporal distribution characteristics of voltage dips and current surges are detected, and the feeder section corresponding to the location of the voltage dip is located using a dynamic topology database; Analyze the changes in power flow direction before and after the fault, and use the upstream and downstream relationships of the feeder section in the dynamic topology database to determine the power interruption point; The continuity of fault paths is verified by combining a dynamic topology database, and the fault propagation path is confirmed by the topological connectivity between the power interruption point and the actual connection graph. Identify the faulty section and isolate the affected area. Based on the fault propagation path, select upstream and downstream operable switches to perform tripping operations.

2. The intelligent protection and switching method capable of self-healing power grid faults according to claim 1, characterized in that: The synchronous acquisition of voltage, current, and power data at each power distribution node also includes: Three-phase electrical quantities are acquired at a millisecond-level frequency using a measurement unit with time synchronization function; The collected data is then linked to a timestamp and uploaded to the regional control center. The voltage drop threshold is set at 70% of the rated voltage, and the current rise rate threshold is five times the rated current per millisecond. Calculate the direction of active power for each line segment and record the changes before and after the fault.

3. The intelligent protection and switching method capable of self-healing from power grid faults according to claim 1, characterized in that: The process of generating the current actual connection diagram by combining switch status information also includes: By overlaying the physical topology with the real-time switch remote signaling status, an actual connectivity diagram reflecting the current power grid connection status is formed. Based on the actual connected graph, traverse all edges to analyze the power flow direction and determine the power convergence point and flow direction. After completing the power flow analysis, nodes with only one power inflow direction are marked as upstream power nodes; Nodes with multiple power input directions are identified as connection points or ring points to distinguish complex network structures.

4. The intelligent protection and switching method capable of self-healing power grid faults according to claim 3, characterized in that: The method of identifying the power supply direction under the current operating mode also includes: By traversing all edges in the actual connected graph and analyzing the power flow towards the convergence point, the single power input source of each node can be determined. If a node has multiple power input directions, it is marked as a tie point or a ring point.

5. The intelligent protection and switching method capable of self-healing from power grid faults according to claim 1, characterized in that: The establishment of a dynamic topology database containing electrical parameters also includes: Write the impedance parameters, rated capacity, and current load rate information of each line into the attribute fields of the corresponding edge in the weighted directed graph; The load type and importance level of each node are written into the attribute field of the corresponding vertex to form a multi-attribute topology model that supports path search and security verification.

6. The intelligent protection and switching method capable of self-healing power grid faults according to claim 1, characterized in that: The spatiotemporal distribution characteristics of the detected voltage dips and current surges also include: Set voltage sag threshold and current rise rate threshold; When the voltage of a node falls below a threshold and is accompanied by a sudden increase in current, the node is determined to be in the fault-affected area. The time and spatial location of such events are recorded to form a preliminary set of suspected fault points.

7. The intelligent protection and switching method capable of self-healing power grid faults according to claim 1, characterized in that: The analysis of the power flow direction changes before and after the fault also includes: Line segments where the power direction was positive before the fault but returned to zero or reversed after the fault were marked as suspected fault paths; Find the shortest connected path between the starting point of the voltage drop and the point of maximum current surge on the suspected fault path. The shortest connected path is used as a candidate fault propagation path for subsequent verification.

8. The intelligent protection and switching method capable of self-healing power grid faults according to claim 1, characterized in that: The process of determining the faulty section and isolating the affected area also includes: Locate the power-side switch and load-side switch closest to the fault point along the fault propagation path; Send a trip command to disconnect the two power supply side switches and the load side switch; The fault is confined to the line segment between the two power-side switches and the load-side switch; The final confirmed fault section is used for subsequent power supply path reconfiguration.

9. The intelligent protection and switching method capable of self-healing power grid faults according to claim 1, characterized in that: The method of verifying fault path continuity using a dynamic topology database also includes: Locate the two nearest operable switches along the fault path: one near the power supply side and the other near the load side. Issue trip commands to both switches to disconnect them, thereby confining the fault to the line segment between them. This line segment is the finally determined fault section.

10. A smart protection and switching system capable of self-healing from power grid faults for implementing the method as described in any one of claims 1-9, characterized in that: include: A measurement unit with time synchronization function is used to synchronously collect voltage, current and power data of each power distribution node; The regional control center is used to generate the current actual connection diagram by combining switch status information, and to verify the energization status of the line segments in the actual connection diagram by referencing voltage, current and power data. The power direction identification unit is used to identify the power supply direction under the current operating mode and determine the power input source of each node based on the verified actual connection diagram. A dynamic topology database is used to store power grid connection information containing electrical parameters and integrate verified actual connection diagrams, power supply direction, line impedance, and capacity information into a weighted directed graph. The fault feature detection unit is used to detect the spatiotemporal distribution characteristics of voltage dips and current surges, and to locate the feeder section corresponding to the location where the voltage dip occurs using a dynamic topology database. The power flow direction analysis unit is used to analyze the change pattern of power flow direction before and after the fault, and to determine the power interruption point by referring to the upstream and downstream relationship of the feeder section in the dynamic topology database. The fault path verification unit is used to verify the continuity of fault paths by combining a dynamic topology database and to confirm the fault propagation path by the topological connectivity between the power interruption point and the actual connection graph. The fault isolation execution unit is used to determine the faulty section and isolate the scope of the fault's impact, and select upstream and downstream operable switches to perform tripping operations based on the fault propagation path.

Images