A method for improving the resilience of a power storage and distribution network based on multi-agent cooperation

Abstract

The application provides a kind of light storage and distribution network flexibility improvement method based on multi-agent cooperation, belongs to the technical field of distribution network optimization, carries out fault diagnosis to the multiple sub-regions after being divided by multi-agent light storage and distribution network area, identifies fault sub-region and non-fault sub-region, carries out initial isolation analysis to the non-fault sub-region adjacent to fault sub-region, obtains initial isolation value, determines initial fault isolation boundary, so as to not only prevent fault diffusion to non-fault area by blocking the connection of fault interference circuit and passive interference circuit, reduce the degree of overall light storage and distribution network being impacted by fault, enhance the resistance flexibility of power grid to fault, but also the agent in fault area can be accurately investigated based on fault circuit information in boundary, the agent in non-fault area can quickly restore power supply support to passive interference circuit, shorten fault repair and system restart time, improve the flexibility of power grid to quickly recover from fault.

Description

Technical Field

[0001] This invention belongs to the field of distribution network optimization technology, specifically a method for improving the resilience of photovoltaic-storage distribution networks based on multi-agent collaboration. Background Technology

[0002] With the increasing penetration rate of new energy sources, photovoltaic-storage distribution networks (integrated energy systems combining photovoltaic power generation, energy storage systems, and distribution networks) have become an important component of new power systems. However, the volatility of distributed power sources in photovoltaic-storage distribution networks (such as the variation of photovoltaic output with sunlight), the dynamic characteristics of energy storage charging and discharging, and the randomness of load demand pose severe challenges to their resilience under fault conditions. After a fault occurs, it is prone to problems such as wide spread, decreased voltage stability, and power outages in non-faulty areas, seriously affecting system reliability.

[0003] In existing technologies, it is difficult to quantify the dynamic interference relationship between faulty and non-faulty circuits, which makes it easy for faults to spread to non-faulty areas through hidden paths, expanding the scope of the fault's impact. Moreover, fault isolation operations often lead to a sharp decrease in the number of power supply lines or a sharp drop in voltage stability in non-faulty areas, causing "power vacuum" or equipment overload, reducing the coordination efficiency of multi-agent systems, and affecting the continuity of power supply. When the isolation operation has a significant impact on the voltage fluctuations and line connectivity in non-faulty areas, it can easily cause equipment protection malfunctions or cascading faults, further weakening the resilience of the power grid.

[0004] Therefore, this invention provides a method for improving the resilience of photovoltaic-storage-distribution networks based on multi-agent collaboration. Summary of the Invention

[0005] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0006] The technical solution adopted by this invention to solve its technical problem is: A method for improving the resilience of photovoltaic-storage-distribution networks based on multi-agent collaboration includes: Step 1: Perform fault diagnosis on the multiple sub-regions after the multi-agent photovoltaic energy storage distribution network area is divided, and identify the faulty sub-regions and non-faulty sub-regions; Step 2: Perform initial isolation analysis on the non-faulty sub-regions adjacent to the faulty sub-region, obtain the fault interference value, obtain the initial fault isolation point, and determine the initial fault isolation boundary; Step 3: Perform isolation impact analysis on each initial fault isolation node on the initial fault isolation boundary, obtain the isolation impact value, evaluate whether the initial fault isolation points within the initial fault isolation boundary are reasonable, and select the fault isolation points to be adjusted. Step 4: Perform spatial location analysis on each adjusted fault isolation point on the initial fault isolation boundary to complete the adjustment operation of the adjusted fault isolation point.

[0007] A further aspect of the present invention is as follows: the process of performing fault diagnosis on multiple sub-regions and identifying faulty and non-faulty sub-regions is as follows: Set a voltage monitoring cycle and divide it into several equal monitoring periods. Obtain the voltage of the sub-region in each monitoring period as the monitoring voltage value for that period. The process of comparing the voltage value monitored during a specific time period with the voltage range monitored during that time period is as follows: If the voltage value monitored during a time period is within the range of the voltage monitored during that time period, then it is a non-faulty sub-region; If the voltage value monitored during a time period does not exist within the range of the voltage monitored during that time period, then it is a fault sub-region.

[0008] A further aspect of this invention is as follows: Initial isolation analysis is performed on the non-faulty sub-regions adjacent to the faulty sub-region. The process for constructing the fault voltage change curve and the non-faulty voltage change curve is as follows: Extract the power supply diagrams from the faulty sub-region and the non-faulty sub-region respectively to obtain the faulty power supply diagram and the non-faulty power supply diagram; Extract multiple faulty power supply trunks within the faulty power supply diagram and multiple non-faulty power supply trunks within the non-faulty power supply diagram; Multiple suspected interference groups are obtained by pairing and combining faulty power supply trunks that are directly electrically connected to the same common boundary node or have an electrically coupled path through the same common boundary node with non-faulty power supply trunks. Within the suspected interference group, the time-period monitoring voltage values ​​of the faulty power supply trunk line during each monitoring period were obtained, as well as the time-period monitoring voltage values ​​of the non-faulty power supply trunk line during each monitoring period. The time-series monitoring voltage values ​​of the faulty power supply trunk line and the time-series monitoring voltage values ​​of the non-faulty power supply trunk line are sorted according to time series and imported into a two-dimensional coordinate system to construct the fault voltage change curve and the non-fault voltage change curve.

[0009] As a further aspect of the present invention, the process for obtaining the fault interference value is as follows: Based on each monitoring period, the fault voltage change curve and the non-fault voltage change curve are divided separately to obtain multiple fault change sub-curves and multiple non-fault change sub-curves. Within the sub-curve analysis group, the endpoint coordinates on the fault change sub-curve and the non-fault change sub-curve are extracted respectively, namely the fault sub-start coordinate, the fault sub-end coordinate, the non-fault sub-start coordinate, and the non-fault sub-end coordinate. The slope is then calculated using the slope calculation formula to obtain the fault sub-slope and the non-fault sub-slope respectively. The fault sub-slope and non-fault sub-slope within each sub-curve analysis group are sequentially input into the Euclidean distance calculation formula according to the time series of each sub-curve analysis group, and the fault interference value is output.

[0010] As a further aspect of the present invention, the process for obtaining the initial fault isolation point is as follows: If the fault interference value is less than or equal to the fault interference threshold, the non-faulty power supply trunk is marked as a passive interference circuit. The connection node is a node in the common boundary node set, or a bus node / segment switch node shared in the topology by the fault interference circuit and the passive interference circuit, which serves as the initial fault isolation node.

[0011] As a further aspect of the present invention, the process for determining the initial fault isolation boundary is as follows: The initial fault isolation nodes are sorted from smallest to largest according to the fault interference values ​​corresponding to each fault interference circuit and passive interference circuit, and the initial fault isolation boundary is determined.

[0012] A further aspect of this invention is as follows: An isolation impact analysis is performed on each initial fault isolation node on the initial fault isolation boundary. The process for obtaining the ratio of affected lines is as follows: Arbitrarily select an initial fault isolation node on the initial fault isolation boundary as the target analysis node. After the isolation operation, count the total number of power supply lines connected to the target analysis node in the non-fault power supply diagram, and use this count as the total number of power supply lines after isolation. The difference between the total number of power supply lines after isolation and the total number of power supply lines before isolation is calculated, and the ratio of the difference to the total number of power supply lines after isolation is calculated to output the ratio of the number of affected lines.

[0013] A further aspect of this invention is as follows: An isolation impact analysis is performed on each initial fault isolation node on the initial fault isolation boundary. The process for obtaining the isolation fluctuation ratio is as follows: Based on the target analysis node, within the non-faulty power supply diagram, select a power supply line connected to the target analysis node as the target line; Obtain the time period monitoring voltage values ​​of the target line for each historical monitoring period before the isolation operation, calculate the standard deviation, and output the line voltage stability value before isolation. Obtain the monitoring voltage values ​​of the target line during each comparative monitoring period after the isolation operation, calculate the standard deviation, and output the stable voltage value of the line after isolation. If the stable voltage value of the line after isolation is greater than the stable voltage value of the line before isolation, it is marked as an isolated fluctuating line. The total number of isolated fluctuating lines is counted and compared with the total number of power supply lines before isolation. The output is the isolated fluctuating line number ratio.

[0014] As a further aspect of the present invention, the process for obtaining the isolation influence value is as follows: The isolation impact value is obtained by summing the ratio of the number of affected lines to the ratio of the number of isolated fluctuations.

[0015] A further aspect of this invention is as follows: spatial location analysis is performed on each adjusted fault isolation point, the process of which is as follows: Starting from the fault isolation point, extend to both sides along the corresponding passive interference circuit, and extract other common boundary nodes between the passive interference circuit and the corresponding fault interference circuit, excluding the current fault isolation point, as candidate adjustment nodes. Based on the order of electrical distance from the current fault isolation point to the nearest and farthest, calculate the isolation impact value corresponding to each candidate adjustment node in turn, and take the first candidate adjustment node whose isolation impact value is less than or equal to the isolation impact threshold as the updated fault isolation point, and terminate the search.

[0016] The beneficial effects of this invention are as follows: 1. This invention performs fault diagnosis on multiple sub-regions after the multi-agent photovoltaic energy storage distribution network is divided, identifies faulty sub-regions and non-faulty sub-regions, performs initial isolation analysis on non-faulty sub-regions adjacent to faulty sub-regions, obtains initial isolation values, and determines initial fault isolation boundaries. This not only prevents the fault from spreading to non-faulty areas by blocking the connection between faulty interference circuits and passive interference circuits, thus reducing the overall impact of faults on the photovoltaic energy storage distribution network and enhancing the resilience of the grid to faults, but also provides an operational framework for multi-agent collaborative recovery based on the clearly defined initial isolation boundaries: agents within the faulty area can accurately troubleshoot based on faulty circuit information within the boundary, while agents in non-faulty areas can quickly restore power supply support to passive interference circuits. This division of labor and collaboration shortens fault repair and system restart time and improves the resilience of the grid to recover quickly from faults. 2. This invention performs isolation impact analysis on each initial fault isolation node on the initial fault isolation boundary, obtaining the isolation impact value. This facilitates the reasonable evaluation of the initial fault isolation point, thereby not only preventing the fault in the fault sub-region from spreading to the adjacent non-fault sub-region, but also reducing the probability of a large-scale power vacuum phenomenon in other non-fault sub-regions after isolation. It assesses whether the initial fault isolation point within the initial fault isolation boundary is reasonable and selects adjustment fault isolation points. This not only reduces the number of power supply lines and voltage stability impact on non-fault areas during fault isolation operations, but also improves the system's anti-disturbance capability, resource allocation efficiency, and power supply continuity when multiple agents coordinate to respond to faults. Furthermore, it performs spatial location analysis on each adjustment fault isolation point on the initial fault isolation boundary, completing the adjustment operation of the adjustment fault isolation point, reducing voltage fluctuation lines and sudden changes in the number of lines, avoiding faults in power equipment in other non-fault sub-regions caused by voltage instability, and reducing the probability of secondary faults. Attached Figure Description

[0017] The invention will now be further described with reference to the accompanying drawings.

[0018] Figure 1 This is a flowchart of the steps of a method for improving the resilience of a photovoltaic-storage-distribution network based on multi-agent collaboration according to the present invention; Figure 2 This is a flowchart of the steps in this invention to evaluate whether the initial fault isolation point within the initial fault isolation boundary is reasonable. Detailed Implementation

[0019] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0020] Example 1 Please see Figure 1 As shown in the figure, the method for improving the resilience of a photovoltaic-storage-distribution network based on multi-agent collaboration according to an embodiment of the present invention includes the following steps: Step 1: Perform fault diagnosis on the multiple sub-regions after the multi-agent photovoltaic energy storage distribution network area is divided, and identify the faulty sub-regions and non-faulty sub-regions; It should be noted that the multi-agent photovoltaic energy storage distribution network area is divided equally according to the grid method to obtain multiple sub-regions, and each sub-region is a regular grid shape, and occupies the same proportion of the total area of ​​the multi-agent photovoltaic energy storage distribution network area. In some embodiments, a voltage monitoring period is set and divided equally into several monitoring periods, with each monitoring period having an equal duration. It should be noted that voltage monitoring operations are performed simultaneously on each sub-area within the multi-agent photovoltaic energy storage distribution network area during the same monitoring period. The voltage of the sub-region during each monitoring period is obtained as the monitoring voltage value for that period. It should be noted that the voltage acquisition rule for each sub-region during each monitoring period is as follows: the voltage is acquired in chronological order within the voltage monitoring cycle of the monitoring period. Therefore, this facilitates timely fault repair and handling of faulty sub-regions. The process of comparing the voltage value monitored during a specific time period with the voltage range monitored during that time period is as follows: If the voltage value monitored during a time period is within the range of the voltage value monitored during that time period, it indicates that no fault occurred in the analyzed sub-region during the monitoring period, and it is a non-faulty sub-region. If the voltage value monitored during a time period is not within the range of the voltage monitored during that time period, it indicates that a fault occurred in the sub-region being analyzed during the monitoring period, and thus it is a faulty sub-region. It should be noted that the voltage range for the specified time period was set by those skilled in the art; Step 2: Perform initial isolation analysis on the non-faulty sub-regions adjacent to the faulty sub-region, obtain the fault interference value, obtain the initial fault isolation point, and determine the initial fault isolation boundary; In some embodiments, power supply maps are extracted from faulty sub-regions and non-faulty sub-regions respectively to obtain faulty power supply maps and non-faulty power supply maps; It should be noted that, from a spatial perspective, the faulty sub-region and the non-faulty sub-region are adjacent sub-regions within the multi-agent photovoltaic energy storage distribution network area; Extract multiple faulty power supply trunks within the faulty power supply diagram and multiple non-faulty power supply trunks within the non-faulty power supply diagram; Based on the distribution network topology, the set of common boundary nodes between faulty sub-regions and non-faulty sub-regions is extracted; among multiple faulty power supply trunks and multiple non-faulty power supply trunks, faulty power supply trunks that are directly electrically connected through the same common boundary node or have an electrical coupling path through the same common boundary node are paired and combined to obtain multiple suspected interference groups; if there is no electrical connection node between faulty power supply trunks and non-faulty power supply trunks, no pairing or combination is performed. It should be noted that the faulty power supply main circuit has an actual electrical connection; Within the suspected interference group, the time-period monitoring voltage values ​​of the faulty power supply trunk line during each monitoring period were obtained, as well as the time-period monitoring voltage values ​​of the non-faulty power supply trunk line during each monitoring period. The time-series monitoring voltage values ​​of the faulty power supply trunk line and the time-series monitoring voltage values ​​of the non-faulty power supply trunk line are sorted according to time series and imported into a two-dimensional coordinate system to construct the fault voltage change curve and the non-faulty voltage change curve. Based on each monitoring period, the fault voltage change curve and the non-fault voltage change curve are divided separately to obtain multiple fault change sub-curves and multiple non-fault change sub-curves. By combining the fault change sub-curve and the non-fault change sub-curve within the same monitoring period into a sub-curve analysis group, multiple sub-curve analysis groups are obtained. Within the sub-curve analysis group, the endpoint coordinates on the fault change sub-curve and the non-fault change sub-curve are extracted respectively, namely the fault sub-start coordinates, the fault sub-end coordinates, the non-fault sub-start coordinates, and the non-fault sub-end coordinates. The coordinates of the starting point and ending point of the fault sub are extracted and connected, and the slope of the fault sub is obtained through the slope calculation formula. Extract the starting coordinates of the non-faulty sub-start and ending coordinates of the non-faulty sub-end, connect them, and obtain the slope of the non-faulty sub-end using the slope calculation formula; The fault sub-slope and non-fault sub-slope within each sub-curve analysis group are sequentially input into the Euclidean distance calculation formula according to the time series of each sub-curve analysis group, and the fault interference value is output. ; The Euclidean distance is calculated as follows: Where r represents the total number of sub-curve analysis groups, Represented as the first Sub-curve analysis of the slope of the corresponding fault sub-group within the group Represented as the first Sub-curve analysis of the slope of the corresponding non-faulty sub-sub within the group; It is understandable that the fault interference value represents an index reflecting the correlation between the voltage fluctuation trends (expressed as slope) of the faulty and non-faulty power supply lines during the same monitoring period. The core logic is that a higher slope similarity (smaller Euclidean distance) indicates a more significant impact of the voltage fluctuations of the faulty line on the non-faulty line, meaning a greater degree of interference. This facilitates the capture of the dynamic voltage correlation between the faulty and non-faulty lines, early detection of hidden interference paths, and solves the problem of difficulty in predicting fault propagation paths. If the fault interference value is greater than the fault interference threshold, it indicates that the voltage fluctuation trend of the faulty power supply trunk and the non-faulty power supply trunk is not very similar during the voltage monitoring period, and the faulty power supply trunk has a small degree of interference to the non-faulty power supply trunk. If the fault interference value is less than or equal to the fault interference threshold, it indicates that the voltage fluctuation trend of the fault power supply trunk and the non-fault power supply trunk is highly similar during the voltage monitoring period, and the fault power supply trunk interferes with the non-fault power supply trunk to a greater extent. The non-fault power supply trunk is then identified as a passive interference circuit. It should be noted that the fault interference threshold is set by those skilled in the art; Extract the connection nodes on each fault interference circuit and passive interference circuit as initial fault isolation nodes, and sort them according to the corresponding fault interference values ​​between each fault interference circuit and passive interference circuit from small to large to determine the initial fault isolation boundary. Among them, the connecting nodes are the nodes in the common boundary node set, or the bus nodes / segment switch nodes shared in the topology by the fault interference circuit and the passive interference circuit. It should be noted that not every two power supply trunks belonging to different sub-regions in a distribution network are physically connected. To avoid invalid analysis of trunks without topological association, a topological validity screening is performed before combining them. The intersection of the set of boundary switches of the faulty sub-region and the set of boundary switches of the non-faulty sub-region is defined as the common boundary node set. Only when a faulty power supply trunk and a non-faulty power supply trunk are simultaneously connected to the same node in the common boundary node set (or form an electrical connection path through the node) are the two combined into a suspected interference group. For trunk road combinations that do not meet the above topology connection conditions, they will not be included in the subsequent calculation of fault interference values. Therefore, each suspected interference group corresponds to at least one clearly defined physical connection node; It should be noted that the purpose of determining the initial fault isolation boundary is: Objective 1: By blocking the connection between the fault interference circuit and the passive interference circuit, the fault is prevented from spreading to the non-faulty area, thereby reducing the overall impact of the fault on the photovoltaic-storage distribution network and enhancing the resilience of the power grid to faults. Objective 2: A clear initial isolation boundary provides an operational framework for multi-agent collaborative recovery: agents within the fault area can accurately troubleshoot based on fault circuit information within the boundary, while agents in non-fault areas can quickly restore power supply support to passive interference circuits. This division of labor and collaboration shortens fault repair and system restart time, and enhances the resilience of the power grid to recover quickly from faults. The specific implementation plan of this embodiment is as follows: Fault diagnosis is performed on multiple sub-regions after the multi-agent photovoltaic energy storage distribution network area is divided, faulty sub-regions and non-faulty sub-regions are identified, and initial isolation analysis is performed on the non-faulty sub-regions adjacent to the faulty sub-regions to obtain the initial isolation value and determine the initial fault isolation boundary. This not only prevents the fault from spreading to the non-faulty area by blocking the connection between the faulty interference circuit and the passive interference circuit, thus reducing the overall impact of the photovoltaic energy storage distribution network on the fault and enhancing the grid's resilience to faults, but also provides an operational framework for multi-agent collaborative recovery by clearly defining the initial isolation boundary: agents in the faulty area can accurately troubleshoot based on the faulty circuit information within the boundary, and agents in the non-faulty area can quickly restore power supply support to the passive interference circuit. This division of labor and collaboration shortens the fault repair and system restart time and improves the resilience of the grid to recover quickly from faults.

[0021] Example 2 Please see Figure 1 As shown in the figure, the method for improving the resilience of a photovoltaic-storage-distribution network based on multi-agent coordination according to an embodiment of the present invention includes the following steps: Step 3: Perform isolation impact analysis on each initial fault isolation node on the initial fault isolation boundary, obtain the isolation impact value, evaluate whether the initial fault isolation points within the initial fault isolation boundary are reasonable, and screen out the fault isolation points to be adjusted. In some embodiments, an initial fault isolation node on the initial fault isolation boundary is extracted as the target analysis node; Based on the target analysis node, extract the power supply line in the non-faulty power supply map corresponding to a non-faulty sub-region; Before the isolation operation, count the total number of power lines connected to the target analysis node in the non-faulty power supply diagram, and use this as the total number of power lines before isolation. After the isolation operation, the total number of power lines connected to the target analysis node in the non-faulty power supply diagram is counted and used as the total number of power lines after isolation. The difference between the total number of power supply lines after isolation and the total number of power supply lines before isolation is calculated, and the ratio of the difference to the total number of power supply lines after isolation is calculated to output the ratio of the number of affected lines. Based on the target analysis node, before the isolation operation, arbitrarily select a power supply line in the non-faulty power supply diagram that is connected to the target analysis node as the target line; After the isolation operation, extract a power supply line within the non-faulty power supply diagram that is connected to the target analysis node, and use it as the target line. It should be noted that the target line and the target line are the same line within the photovoltaic-storage distribution network. The only difference is that the target line was selected before the isolation operation and the target line was selected after the isolation operation. Extract the historical monitoring period, divide the historical monitoring period into several equal historical monitoring periods, and ensure that each historical monitoring period has an equal duration; It should be noted that the historical monitoring period is the total monitoring process duration corresponding to the voltage monitoring operation of the target line; For the target line, the monitoring voltage value for each historical monitoring period is obtained, the standard deviation is calculated, and the voltage stability value of the analyzed line is output. Similarly, a comparison monitoring cycle is set, and the comparison monitoring cycle is divided into several comparison monitoring periods of equal duration. The method of dividing the comparison monitoring period is consistent with the method of dividing the historical monitoring period, and the duration of each comparison monitoring period is equal to the duration of the historical monitoring period. For the target line, the monitoring voltage value for each time period corresponding to the comparison monitoring period is obtained, and the standard deviation is calculated to output the voltage stability value of the comparison line. The comparison between the stable line voltage values ​​and the analyzed line voltage stability values ​​is performed as follows: If the voltage stability value of the comparison line is greater than the voltage stability value of the analysis line, it indicates that after the isolation operation, the voltage of the power supply line connected to the target analysis node in the photovoltaic-storage distribution network is more unstable than the voltage of the power supply line connected to the target analysis node in the photovoltaic-storage distribution network before the isolation operation, and is identified as the isolated fluctuating line. If the voltage stability value of the compared line is less than or equal to the voltage stability value of the analyzed line, it indicates that after the isolation operation, the voltage of the power supply line connected to the target analysis node in the photovoltaic-storage distribution network is more stable than the voltage change of the power supply line connected to the target analysis node in the photovoltaic-storage distribution network before the isolation operation, and is identified as an isolated stable line. Count the total number of isolated fluctuation lines and compare it with the total number of power supply lines before isolation to obtain the ratio of isolated fluctuation lines. The sum of the ratio of affected lines to the ratio of isolated fluctuations is used to output the isolation impact value; It can be understood that the meaning of the isolation impact value is: by quantitatively analyzing the changes in the power supply lines of the initial fault isolation node in the photovoltaic-storage distribution network before and after the isolation operation, the changes in the power supply lines include the connection and voltage stability of the power supply lines of the target analysis node to the non-fault area, which is conducive to the reasonable evaluation of the initial fault isolation point. This not only avoids the spread of the fault in the fault sub-region to the adjacent non-fault sub-region, but also reduces the probability of a large-scale power vacuum phenomenon in other non-fault sub-regions after isolation. The isolation impact value is compared with the isolation impact threshold, as follows: If the isolation impact value is greater than the isolation impact threshold, it indicates that the target analysis node has a significant impact on the power supply line connectivity and voltage stability of the non-faulty area. A fault isolation point adjustment signal is generated and marked as an adjustment fault isolation point. If the isolation impact value is less than or equal to the isolation impact threshold, it indicates that the target analysis node has a small impact on the power supply line connectivity and voltage stability of the non-faulty area, and a reasonable fault isolation point signal is generated. It is understandable that the threshold for isolation impact is set by those skilled in the art; It should be noted that the purpose of adjusting the fault isolation point is: Objective 1: By optimizing the node selection of the isolation boundary, reduce the number of power supply lines and voltage stability impact on non-faulty areas during fault isolation operations, thereby improving the system's anti-disturbance capability, resource allocation efficiency, and power supply continuity when multiple agents work together to deal with faults. Objective 2: To reduce voltage fluctuations and sudden changes in the number of lines, thereby preventing power equipment failures in other non-faulty sub-areas due to voltage instability and reducing the probability of secondary failures; Step 4: Perform spatial location analysis on each adjusted fault isolation point on the initial fault isolation boundary to complete the adjustment operation of the adjusted fault isolation point; In some embodiments, spatial location analysis is performed on each adjusted fault isolation point, as follows: Starting from the current fault isolation point, obtain the passive interference circuit and fault interference circuit corresponding to the location of the current fault isolation point; Based on the distribution network topology adjacency relationship, the passive interference circuit is searched outward step by step along the extension direction of the passive interference circuit. All common boundary nodes between the passive interference circuit and the corresponding fault interference circuit, except for the current adjustment fault isolation point (i.e., the segmented switch nodes or tie switch nodes that connect two trunk lines at the boundary between the fault sub-region and the non-fault sub-region), are extracted and defined as the candidate adjustment node set. Obtain the electrical distance (i.e., the number of switches / buses along the power supply line, or the per-unit value of the line impedance) between each node in the candidate adjustment node set and the current adjustment fault isolation point, and sort the candidate adjustment nodes in order of electrical distance from near to far to determine the search priority; At the same time, according to the isolation impact value corresponding to each candidate adjustment node, the first candidate adjustment node that meets the condition that the isolation impact value is less than or equal to the isolation impact threshold is taken as the updated fault isolation point, and the search is terminated immediately to complete the adjustment operation of the fault isolation point. After traversing all nodes in the candidate adjustment node set, if no node has an isolation impact value less than or equal to the isolation impact threshold, it means that the impact requirements cannot be met by adjusting this node under the current topology constraints. In this case, the current adjustment failure isolation point remains unchanged, no physical switching action is performed, and a 'boundary optimization failure prompt message' is generated and output to the upper-level scheduling module of the multi-agent collaborative system for manual review by the scheduler or to trigger a global isolation strategy recalculation.

[0022] The specific solution of this invention is as follows: An isolation impact analysis is performed on each initial fault isolation node on the initial fault isolation boundary to obtain the isolation impact value. This facilitates a reasonable evaluation of the initial fault isolation point, thereby not only preventing the fault in the fault sub-region from spreading to adjacent non-fault sub-regions, but also reducing the probability of a large-scale power vacuum phenomenon in other non-fault sub-regions after isolation. The rationality of the initial fault isolation point within the initial fault isolation boundary is evaluated, and adjustment fault isolation points are selected. This not only reduces the number of power supply lines and voltage stability impact on non-fault areas during fault isolation operations, but also improves the system's anti-disturbance capability, resource allocation efficiency, and power supply continuity when multiple agents coordinate to respond to faults. Spatial location analysis is performed on each adjustment fault isolation point on the initial fault isolation boundary to complete the adjustment operation, reducing voltage fluctuation lines and sudden changes in the number of lines, avoiding faults in power equipment in other non-fault sub-regions caused by voltage instability, and reducing the probability of secondary faults.

[0023] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for improving the resilience of photovoltaic-storage-distribution networks based on multi-agent collaboration, characterized in that: include: Step 1: Perform fault diagnosis on the multiple sub-regions after the multi-agent photovoltaic energy storage distribution network area is divided, and identify the faulty sub-regions and non-faulty sub-regions; Step 2: Perform initial isolation analysis on the non-faulty sub-regions adjacent to the faulty sub-region, obtain the fault interference value, obtain the initial fault isolation point, and determine the initial fault isolation boundary; Step 3: Perform isolation impact analysis on each initial fault isolation node on the initial fault isolation boundary, obtain the ratio of the number of affected lines and the ratio of the number of isolated fluctuations, process them, evaluate whether the initial fault isolation points within the initial fault isolation boundary are reasonable, and screen out the fault isolation points to be adjusted. Step 4: Perform spatial location analysis on each adjusted fault isolation point on the initial fault isolation boundary to complete the adjustment operation of the adjusted fault isolation point.

2. The method for improving the resilience of photovoltaic-storage-distribution networks based on multi-agent collaboration as described in claim 1, characterized in that: The process of fault diagnosis for multiple sub-regions, identifying faulty and non-faulty sub-regions, is as follows: Set a voltage monitoring cycle and divide it into several equal monitoring periods. Obtain the voltage of the sub-region in each monitoring period as the monitoring voltage value for that period. The process of comparing the voltage value monitored during a specific time period with the voltage range monitored during that time period is as follows: If the voltage value monitored during a time period is within the range of the voltage monitored during that time period, then it is a non-faulty sub-region; If the voltage value monitored during a time period does not exist within the range of the voltage monitored during that time period, then it is a fault sub-region.

3. The method for improving the resilience of photovoltaic-storage-distribution networks based on multi-agent collaboration according to claim 1, characterized in that: Initial isolation analysis is performed on the non-faulty sub-regions adjacent to the faulty sub-region. The process of constructing the fault voltage change curve and the non-faulty voltage change curve is as follows: Extract the power supply diagrams from the faulty sub-region and the non-faulty sub-region respectively to obtain the faulty power supply diagram and the non-faulty power supply diagram; Extract multiple faulty power supply trunks within the faulty power supply diagram and multiple non-faulty power supply trunks within the non-faulty power supply diagram; Multiple suspected interference groups are obtained by pairing and combining faulty power supply trunks that are directly electrically connected to the same common boundary node or have an electrically coupled path through the same common boundary node with non-faulty power supply trunks. Within the suspected interference group, the time-period monitoring voltage values ​​of the faulty power supply trunk line during each monitoring period were obtained, as well as the time-period monitoring voltage values ​​of the non-faulty power supply trunk line during each monitoring period. The time-series monitoring voltage values ​​of the faulty power supply trunk line and the time-series monitoring voltage values ​​of the non-faulty power supply trunk line are sorted according to time series and imported into a two-dimensional coordinate system to construct the fault voltage change curve and the non-fault voltage change curve.

4. The method for improving the resilience of photovoltaic-storage-distribution networks based on multi-agent collaboration as described in claim 1, characterized in that: The process of obtaining fault interference values ​​is as follows: Based on each monitoring period, the fault voltage change curve and the non-fault voltage change curve are divided separately to obtain multiple fault change sub-curves and multiple non-fault change sub-curves. Within the sub-curve analysis group, the endpoint coordinates on the fault change sub-curve and the non-fault change sub-curve are extracted respectively, namely the fault sub-start coordinate, the fault sub-end coordinate, the non-fault sub-start coordinate, and the non-fault sub-end coordinate. The slope is then calculated using the slope calculation formula to obtain the fault sub-slope and the non-fault sub-slope respectively. The fault sub-slope and non-fault sub-slope within each sub-curve analysis group are sequentially input into the Euclidean distance calculation formula according to the time series of each sub-curve analysis group, and the fault interference value is output.

5. The method for improving the resilience of photovoltaic-storage-distribution networks based on multi-agent collaboration according to claim 1, characterized in that: The process of obtaining the initial fault isolation point is as follows: If the fault interference value is less than or equal to the fault interference threshold, the non-faulty power supply trunk is marked as a passive interference circuit. Extract the connection nodes on each faulty interference circuit and passive interference circuit as initial fault isolation nodes.

6. The method for improving the resilience of a photovoltaic-storage-distribution network based on multi-agent collaboration as described in claim 5, characterized in that: The process for determining the initial fault isolation boundary is as follows: Based on the magnitude of the fault interference value, the initial fault isolation nodes are sorted in ascending order of the fault interference value to determine the initial fault isolation boundary.

7. The method for improving the resilience of photovoltaic-storage-distribution networks based on multi-agent collaboration according to claim 1, characterized in that: An isolation impact analysis is performed on each initial fault isolation node on the initial fault isolation boundary. The process for obtaining the ratio of affected lines is as follows: Arbitrarily select an initial fault isolation node on the initial fault isolation boundary as the target analysis node. After the isolation operation, count the total number of power supply lines connected to the target analysis node in the non-fault power supply diagram, and use this count as the total number of power supply lines after isolation. The difference between the total number of power supply lines after isolation and the total number of power supply lines before isolation is calculated, and the ratio of the difference to the total number of power supply lines after isolation is calculated to output the ratio of the number of affected lines.

8. The method for improving the resilience of photovoltaic-storage-distribution networks based on multi-agent collaboration according to claim 1, characterized in that: An isolation impact analysis is performed on each initial fault isolation node on the initial fault isolation boundary. The process for obtaining the isolation fluctuation ratio is as follows: Based on the target analysis node, within the non-faulty power supply diagram, select a power supply line connected to the target analysis node as the target line; Obtain the time period monitoring voltage values ​​of the target line for each historical monitoring period before the isolation operation, calculate the standard deviation, and output the line voltage stability value before isolation. Obtain the monitoring voltage values ​​of the target line during each comparative monitoring period after the isolation operation, calculate the standard deviation, and output the stable voltage value of the line after isolation. If the stable voltage value of the line after isolation is greater than the stable voltage value of the line before isolation, it is marked as an isolated fluctuating line. The total number of isolated fluctuating lines is counted and compared with the total number of power supply lines before isolation. The output is the isolated fluctuating line number ratio.

9. The method for improving the resilience of photovoltaic-storage-distribution networks based on multi-agent collaboration according to claim 1, characterized in that: The process for adjusting the selection of fault isolation points is as follows: The sum of the ratio of affected lines to the ratio of isolated fluctuations is used to output the isolation impact value; If the isolation impact value is greater than the isolation impact threshold, a fault isolation point adjustment signal is generated and marked as an adjustment fault isolation point.

10. A method for improving the resilience of a photovoltaic-storage-distribution network based on multi-agent collaboration as described in claim 1, characterized in that: Spatial location analysis was performed on each adjusted fault isolation point, as follows: Starting from the fault isolation point, extend to both sides along the corresponding passive interference circuit, and extract other common boundary nodes between the passive interference circuit and the corresponding fault interference circuit, excluding the current fault isolation point, as candidate adjustment nodes. Based on the order of electrical distance from the current fault isolation point to the nearest and farthest, calculate the isolation impact value corresponding to each candidate adjustment node in turn, and take the first candidate adjustment node whose isolation impact value is less than or equal to the isolation impact threshold as the updated fault isolation point, and terminate the search.

Images