Self-healing control system and control method based on multi-agent cooperation

Abstract

The invention relates to the field of power distribution control, and particularly discloses a self-healing control system and method based on multi-agent collaboration, and the system comprises a regional agent and a local agent which are disposed in a power distribution network. Wherein the local intelligent agent is configured to obtain real-time operation data of a monitoring point in the power distribution network, and perform rapid fault detection on the real-time operation data; the regional intelligent agent is configured to fuse the fault analysis information of the plurality of local intelligent agents to determine fault position information when the power distribution network has a fault, generate a self-healing control strategy according to the fault position information, and send the self-healing control strategy to the local intelligent agents for execution; through cooperative work of the distributed intelligent agents, the problems of low response speed, low positioning precision, weak autonomous capability and the like in traditional power distribution network fault management are solved. In addition, the system does not need to depend on centralized control of a master station, can still operate autonomously when communication is interrupted, and improves the efficiency and reliability of fault processing through cooperation of multiple agents.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of power distribution control, and particularly to a self-healing control system and method based on multi-agent collaboration. BACKGROUND

[0002] Traditional power distribution network fault management mainly relies on manual inspection and SCADA (Supervisory Control and Data Acquisition) systems. However, with the increasing complexity of modern power distribution network structure and the high proportion of distributed generation (DG) access, these traditional methods face severe challenges.

[0003] Firstly, slow fault response speed is the core defect of current technology. Traditional overcurrent protection needs hundreds of milliseconds to respond, which is far from meeting the growing demand for fast fault isolation of power systems. In the case of a large number of new energy access, any delay may lead to a chain reaction, affecting the stability of the entire power grid.

[0004] Secondly, low fault positioning accuracy is another big problem. The traditional impedance method is easily affected by the access of distributed power, resulting in positioning error exceeding 500 meters, which makes fault investigation and recovery work inefficient.

[0005] Finally, the existing system has weak autonomy and relies heavily on centralized control of the master station. When the communication link in the power distribution network is interrupted, the master station cannot issue control instructions, resulting in the system being unable to autonomously isolate faults and restore power supply. SUMMARY

[0006] The purpose of the present application is to provide a self-healing control system and method based on multi-agent collaboration, which adopts distributed intelligent agents to work collaboratively, improving the efficiency and reliability of fault handling.

[0007] The present application discloses a self-healing control system based on multi-agent collaboration, comprising: regional agents and local agents arranged in a power distribution network; The local agent is configured to acquire real-time operation data of monitoring points in the power distribution network and perform rapid fault detection on the real-time operation data; The regional agent is configured to fuse fault analysis information of multiple local agents when the power distribution network fails, determine fault location information, and generate a self-healing control strategy according to the fault location information; The local agent is also configured to execute the self-healing control strategy issued by the regional agent on the power distribution network when the power distribution network fails.

[0008] Further, the local agent comprises: The fault detection module is configured to detect a sudden change of voltage or current in the power distribution network in real time based on a PMU phasor difference method, analyze the sudden change of voltage or current based on a wavelet transform algorithm, and extract a fault feature of the voltage or current; The fault feature uploading module is configured to upload the fault feature to the regional agent.

[0009] Further, the regional agent comprises: The fault analysis module is configured to locate and analyze the fault feature. The fault location module is configured to determine fault location information based on the fault feature. The self-healing control strategy generation module is configured to determine the self-healing control strategy based on the fault location information.

[0010] Further, the fault analysis module comprises: The topology construction module is configured to construct a power distribution network topology graph. The fault current analysis module is configured to analyze a fault current direction on the power distribution network topology graph based on the graph neural network, and determine a candidate fault region. The interference elimination module is configured to fuse data of feeder terminal units, and verify the candidate fault region. The fault distance determination unit is configured to compare voltage or current phasor changes of each node of the power distribution network before and after a fault, and calculate a fault distance based on an impedance matrix. The traveling wave positioning unit is configured to analyze a traveling wave generated when a fault occurs.

[0011] Further, the fault location module comprises: The information exchange module is configured to control a plurality of regional agents to exchange boundary measurement values of power distribution networks under their jurisdiction with each other. The step-by-step algorithm optimization unit is configured to iteratively solve a collaborative positioning result based on the boundary measurement values by using an alternating direction multiplier algorithm. The agent voting unit is configured to determine a fault location by using a consensus algorithm in combination with the collaborative positioning result, the fault distance, and traveling wave positioning information. The agent voting compensation unit dynamically allocates voting weights of the regional agents based on PMU data reliability.

[0012] Further, the self-healing control strategy generation module comprises: The path searching unit is configured to search for an optimal power supply recovery path based on the power distribution network topology graph. The islanding unit is configured to generate a minimum islanding scheme containing a sequence of switching actions based on the optimal power restoration path. The resource allocation unit is configured to verify the minimum islanding scheme. If the distribution network is overloaded after adopting the minimum islanding scheme, a power balance scheme is generated through a smart negotiation algorithm. The issuing unit is configured to issue the minimum islanding scheme generated by the islanding unit, or to issue the minimum islanding scheme after power balancing adjustment.

[0013] Furthermore, the regional intelligent agent also includes: The adaptive control unit is configured to use a model predictive control algorithm to optimize the control action of the smart switch in the next few seconds.

[0014] Furthermore, the system also includes a central decision-making agent configured to enable multiple regional agents to work collaboratively.

[0015] On the other hand, the present invention also discloses a self-healing control method based on multi-agent cooperation, the method comprising: Set up regional intelligent agents and local intelligent agents in the power distribution network; The local intelligent agent is used to acquire real-time operating data of monitoring points in the power distribution network and to perform rapid fault detection on the real-time operating data. When the power distribution network fails, the regional agent integrates the fault analysis information of multiple local agents to determine the fault location information, and generates a self-healing control strategy based on the fault location information. When the distribution network fails, the local agent executes the self-healing control strategy issued by the regional agent on the distribution network.

[0016] Compared with the prior art, the present invention has at least the following technical effects: This invention addresses the problems of slow response speed, low positioning accuracy, and weak autonomy in traditional power distribution network fault management through the collaborative work of distributed intelligent agents. Specifically, local intelligent agents achieve rapid fault detection, regional intelligent agents improve positioning accuracy by fusing multi-source information, and autonomous control is achieved through distributed decision-making. This system does not rely on centralized control from a master station, can still operate autonomously during communication interruptions, and improves the efficiency and reliability of fault handling through multi-agent collaboration. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the self-healing control method in Embodiment 2 of the present invention. Detailed Implementation

[0018] The following detailed description, with reference to schematic diagrams, outlines a self-healing control system and method based on multi-agent collaboration, illustrating preferred embodiments of the invention. It should be noted that the following description is intended to enable those skilled in the art to understand and implement the invention, and is illustrative rather than limiting its scope. Those skilled in the art should understand that, based on the teachings of this specification, various modifications, equivalent substitutions, or improvements can be made to the described specific embodiments without departing from the core spirit and principles of the invention. For example, technical features from different embodiments can be cross-combined to form new technical solutions. As long as no technical contradictions arise, these variations and combinations should fall within the scope of protection claimed by the present invention.

[0019] The invention is described more specifically by way of example in the following paragraphs with reference to the accompanying drawings. The advantages and features of the invention will become clearer from the following description. It should be noted that the drawings are in a very simplified form and use non-precise proportions, and are only used to facilitate and clarify the illustration of the embodiments of the invention.

[0020] Example 1 This embodiment provides a self-healing control system based on multi-agent cooperation. The system includes a regional agent and a local agent located in the power distribution network.

[0021] The local intelligent agent is configured to acquire real-time operating data of monitoring points in the power distribution network and perform rapid fault detection on the real-time operating data.

[0022] The regional agent is configured to, when the power distribution network fails, integrate fault analysis information from multiple local agents to determine the fault location information and generate a self-healing control strategy based on the fault location information.

[0023] The local agent is also configured to execute a self-healing control strategy issued by the regional agent on the distribution network when the distribution network fails.

[0024] In this embodiment, the collaborative work of distributed intelligent agents solves the problems of slow response speed, low positioning accuracy, and weak autonomy in traditional power distribution network fault management. Specifically, the local intelligent agent achieves rapid fault detection, the regional intelligent agent improves positioning accuracy by fusing multi-source information, and autonomous control is achieved through distributed decision-making. This system does not rely on centralized control from a master station, can still operate autonomously during communication interruptions, and improves the efficiency and reliability of fault handling through multi-agent collaboration.

[0025] Furthermore, the local intelligent agent includes: The fault detection module is configured to detect sudden changes in voltage or current in the distribution network in real time based on the phasor difference method of PMU (phasor measurement unit), analyze the sudden changes in voltage or current based on wavelet transform algorithm, and extract the fault characteristics of voltage or current.

[0026] The fault feature uploading module is configured to upload the fault features to the regional smart body.

[0027] Specifically, the PMU phasor difference method detects microsecond-level abrupt changes by comparing the changes in voltage or current phasor amplitude and phase angle between adjacent sampling periods. Wavelet transform can be used to reduce computational complexity, and feature extraction supports simultaneous output of transient component energy distribution and abrupt change timing calibration.

[0028] In a specific example, the phasor difference threshold can be set to a nominal value of 5%-10%.

[0029] In another specific example, the wavelet transform algorithm can use the db4 wavelet basis function for 6-level decomposition, detect singularities by modulus maxima, and extract fault feature frequency components concentrated in the range of 0.1-2kHz.

[0030] In another specific example, an FPGA chip can be integrated into the fault detection module to achieve fast sampling at the 1μs level, and a sliding time window algorithm can be used to eliminate noise interference.

[0031] In another specific example, the fault feature uploading module adopts the IEC 61850-9-2 standard communication protocol, and the transmission delay is controlled within 20ms.

[0032] Therefore, this technical solution achieves millisecond-level detection and accurate extraction of fault characteristics in distribution networks through the coordinated processing of PMU phasor difference method and wavelet transform. Compared with the threshold comparison method used in traditional SCADA systems, this solution improves fault detection sensitivity by an order of magnitude, enabling the identification of voltage sags below 5%. The multi-resolution analysis of wavelet transform effectively solves the problem of mixed fault characteristics caused by distributed power source access, and the feature localization error is controlled within one sampling interval. The standardized communication protocol of the fault feature uploading module enables interconnection between devices from different manufacturers, providing a high-quality data foundation for subsequent collaborative localization of regional intelligent agents.

[0033] Furthermore, the regional intelligent agent includes: The fault analysis module is configured to locate the fault characteristics and analyze them.

[0034] Therefore, the positioning module is configured to determine the fault location information based on the fault characteristics.

[0035] The self-healing control strategy generation module is configured to determine the self-healing control strategy based on the fault location information.

[0036] Furthermore, in this embodiment, the fault analysis module includes a topology construction module, a fault current analysis module, an interference cancellation module, a fault distance determination unit, and a traveling wave positioning unit.

[0037] The topology building module is used to create a distribution network topology diagram.

[0038] In a specific example, the topology graph may be stored in the form of a graph database or an adjacency matrix.

[0039] The fault current analysis module is configured to analyze the direction of fault current on the distribution network topology map based on the graph neural network to determine candidate fault regions.

[0040] In a specific example, a graph neural network can be implemented using architectures such as GraphSAGE or GAT.

[0041] The interference cancellation module is configured to integrate data from the feeder terminal unit to verify the candidate fault area.

[0042] In a specific example, the interference cancellation module may employ a Kalman filter algorithm to eliminate measurement noise.

[0043] The fault distance determination unit is configured to compare the voltage or current phasor changes of each node in the distribution network before and after the fault, and calculate the fault distance based on the impedance matrix.

[0044] In a specific example, the impedance matrix can be obtained by inverting the nodal admittance matrix.

[0045] The traveling wave localization unit is configured to analyze the traveling wave generated when a fault occurs.

[0046] In a specific example, the traveling wave positioning unit can use wavelet transform to extract the arrival time difference of the traveling wave.

[0047] Specifically, the fault analysis module achieves precise location through multi-dimensional data fusion. The topology construction module establishes a power grid structure model, providing a basic framework for subsequent analysis. The fault current analysis module uses graph neural networks to capture the propagation path of fault currents in the power grid, effectively identifying candidate fault areas. The interference cancellation module eliminates false positives through data verification, improving location reliability. The fault distance determination unit combines the impedance method principle with phasor comparison to achieve meter-level accuracy in location. The traveling wave location unit uses high-frequency transient signals to achieve nanosecond-level time synchronization, complementing the impedance method. Cross-validation of multiple location results significantly improves location accuracy in complex distribution network environments.

[0048] Therefore, this embodiment achieves rapid and accurate identification of distribution network faults by establishing a power grid topology model and combining various location methods such as graph neural network analysis, multi-source data verification, impedance method, and traveling wave method. It effectively solves the problem of large location errors caused by distributed power sources in the traditional impedance method, and improves the location reliability in complex power grid environments through multi-method collaboration. The modules in the technical solution work collaboratively, enabling the autonomous completion of the entire process from fault detection to location without relying on centralized control from a master station.

[0049] Furthermore, the fault location module includes: The information exchange module is configured to control multiple regional agents to exchange boundary measurement values ​​of the distribution networks under their jurisdiction.

[0050] The step-by-step algorithm optimization unit is configured to perform iterative calculations based on boundary measurements using an alternating direction multiplier algorithm to solve for the cooperative positioning results.

[0051] The intelligent agent voting unit is configured to combine the cooperative localization results, the fault distance, and the traveling wave localization information to determine the fault location through a consensus algorithm.

[0052] The agent voting compensation unit dynamically allocates the voting weights of the regional agents based on the credibility of the PMU data.

[0053] Specifically, the information exchange module achieves data synchronization among regional agents through a high-speed communication network, where boundary measurements include voltage amplitude, phase angle, and current amplitude. When the distributed algorithm optimization unit employs the alternating direction multiplier algorithm, the fault location problem is modeled as a distributed optimization problem, and parallel computation is achieved through decomposition and coordination methods. When the agent voting unit uses a consensus algorithm, regional agents reach consensus through multiple rounds of information exchange. The agent voting compensation unit dynamically adjusts voting weights based on PMU data quality indicators, such as signal-to-noise ratio and sampling accuracy.

[0054] Therefore, this embodiment effectively solves the failure problem of traditional centralized positioning methods when communication is interrupted through a distributed cooperative positioning mechanism. The application of the alternating direction multiplier algorithm improves computational efficiency, enabling fault location in large-scale distribution networks to be completed within seconds. The dynamic weight allocation mechanism allows high-precision PMU data to play a dominant role in positioning decisions, thereby improving the reliability of positioning results. Compared with the traditional impedance method, this method controls the positioning error within 50 meters, significantly improving the efficiency of fault handling.

[0055] Furthermore, the self-healing control strategy generation module includes: The path search unit is configured to search for the optimal power supply restoration path based on the power distribution network topology.

[0056] The islanding unit is configured to generate a minimum islanding scheme containing a sequence of switching actions based on the optimal power restoration path.

[0057] The resource allocation unit is configured to verify the minimum islanding scheme. If the distribution network is overloaded after adopting the minimum islanding scheme, a power balance scheme is generated through intelligent negotiation algorithm.

[0058] The issuing unit is configured to issue the minimum islanding scheme generated by the islanding unit, or to issue the minimum islanding scheme after power balancing adjustment.

[0059] The path search unit is configured to search for the optimal power restoration path based on the distribution network topology. The islanding unit is configured to generate a minimum islanding scheme containing a sequence of switching actions based on the optimal power restoration path. The resource allocation unit is configured to verify the minimum islanding scheme; if the distribution network becomes overloaded after adopting the minimum islanding scheme, a power balancing scheme is generated through a smart negotiation algorithm. The distribution unit is configured to distribute the minimum islanding scheme generated by the islanding unit, or the minimum islanding scheme adjusted for power balancing.

[0060] In a specific example, the path search unit can use Dijkstra's algorithm or the A* algorithm to find the optimal path, where the path weight can take into account factors such as line impedance, switch status and load priority.

[0061] In another specific example, during the generation of island partitioning schemes, a mixed-integer linear programming method can be used to solve for the minimum combination of switching actions, while simultaneously satisfying the requirements of maximizing topological connectivity and power restoration range.

[0062] In another specific example, the intelligent negotiation algorithm in the resource allocation unit can adopt a game theory-based distributed optimization method to achieve power balance adjustment through information interaction between local agents.

[0063] In this embodiment, a hierarchical collaborative decision-making mechanism is used to achieve rapid self-healing after a distribution network fault. First, topology analysis is used to determine the optimal recovery path. Then, an optimization algorithm generates feasible islanding schemes. Finally, a distributed negotiation mechanism is employed to address potential power imbalances. Compared to traditional centralized control methods, this scheme offers higher decision-making efficiency and fault tolerance, and can maintain basic self-healing capabilities even in abnormal situations such as communication interruptions.

[0064] Furthermore, the regional intelligent agent also includes an adaptive control unit configured to optimize the control actions of the intelligent switch in the next few seconds using a model predictive control algorithm.

[0065] In a specific example, the adaptive control unit first collects real-time status data of the smart switch, including switch position and current and voltage measurements; then it constructs a prediction model based on dynamic power flow equations, which takes into account the output fluctuations of distributed power sources and load changes; finally, it generates the optimal switching action sequence for the next 3-5 seconds by solving a constrained optimization problem.

[0066] In another specific example, the adaptive control unit may employ a distributed solution method to decompose the global optimization problem into multiple sub-problems, which are then solved in parallel by agents in different regions.

[0067] In this embodiment, model predictive control is used to achieve forward-looking optimization, which effectively solves the problem of response lag in traditional control methods. This allows for the early prediction of system state changes and the preparation of control strategies before a fault occurs. The rolling optimization mechanism continuously corrects control commands to adapt to dynamic changes in the system. Long-term control problems are decomposed into multiple short-term optimizations, reducing computational complexity.

[0068] In this embodiment, the local agent is used for data acquisition, data extraction, and control reset, while the regional agent is used for communication and algorithm operation. The local agent is used for rapid on-site data acquisition, extraction, and control reset to achieve faster physical response. Simultaneously, the regional agent centrally handles communication, algorithm operation, and data analysis, constructing a high-fidelity raw data lake by aggregating all local data. This data lake facilitates offline analysis, model training, and deep post-analysis, solving the problems of data dispersion and inconsistent standards in traditional architectures.

[0069] Furthermore, the system also includes a central decision-making agent configured to enable multiple regional agents to work collaboratively.

[0070] In this embodiment, the regional intelligent agent further includes a communication module. The communication module is configured to communicate with the local intelligent agent.

[0071] In one specific example, the communication module is configured to exchange data with the local agent using the IEC61850 GOOSE / SV or IEEEC37.118 protocol to achieve low-latency communication.

[0072] Furthermore, the system also includes a central decision-making agent configured to enable multiple regional agents to work collaboratively.

[0073] In a specific example, the central decision-making agent can be deployed on a cloud platform or edge computing node and communicate with regional agents through message queues or publish / subscribe mechanisms.

[0074] In another specific example, the central decision-making agent may include the following functional units: The collaborative scheduling unit is responsible for receiving the status information of agents in each region and generating global optimization instructions.

[0075] The conflict resolution unit is used to handle control policy conflicts between regional agents.

[0076] The fault-tolerant management unit is used to detect the operating status of the regional intelligent agent and activate the backup mechanism in case of an anomaly.

[0077] This technical solution effectively addresses the decision-making conflicts and inefficiencies that may arise when multiple regional agents collaborate. By introducing a central decision-making agent, the central agent can integrate operational data from various regions and generate optimal coordination strategies based on a global perspective, thereby improving the overall efficiency of fault location and self-healing control.

[0078] Example 2 Based on the same inventive concept, this embodiment proposes a self-healing control method based on multi-agent cooperation, which includes the following steps: S1. Set up regional intelligent agents and local intelligent agents in the distribution network; S2. The local intelligent agent is used to acquire real-time operating data of monitoring points in the power distribution network, and the real-time operating data is used for rapid fault detection; S3. When the power distribution network fails, the regional agent integrates the fault analysis information of multiple local agents to determine the fault location information, and generates a self-healing control strategy based on the fault location information. S4. When the distribution network fails, the local agent executes the self-healing control strategy issued by the regional agent on the distribution network.

[0079] It is understandable that the technical effects that can be achieved by the above-mentioned self-healing control method based on multi-agent collaboration can also be achieved by the self-healing control system based on multi-agent collaboration disclosed in Embodiment 1, so they will not be repeated here.

[0080] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A self-healing control system based on multi-agent cooperation, characterized in that, include: Regional intelligent agents and local intelligent agents located in the power distribution network; The local intelligent agent is configured to acquire real-time operating data of monitoring points in the power distribution network and perform rapid fault detection on the real-time operating data; The regional agent is configured to, when the power distribution network fails, integrate fault analysis information from multiple local agents to determine the fault location information and generate a self-healing control strategy based on the fault location information. The local agent is also configured to execute a self-healing control strategy issued by the regional agent on the distribution network when the distribution network fails.

2. The self-healing control system based on multi-agent collaboration as described in claim 1, characterized in that, The local intelligent agent includes: The fault detection module is configured to detect the sudden changes in voltage or current in the distribution network in real time based on the PMU phasor difference method, analyze the sudden changes in voltage or current based on the wavelet transform algorithm, and extract the fault characteristics of the voltage or current. The fault feature uploading module is configured to upload the fault features to the regional smart body.

3. The self-healing control system based on multi-agent cooperation as described in claim 2, characterized in that, The regional intelligent agent includes: The fault analysis module is configured to locate the fault characteristics and analyze the fault characteristics; Therefore, the fault location module is configured to determine the fault location information based on the fault characteristics; The self-healing control strategy generation module is configured to determine the self-healing control strategy based on the fault location information.

4. The self-healing control system based on multi-agent cooperation as described in claim 3, characterized in that, The fault analysis module includes: The topology building module is configured to build a distribution network topology diagram; The fault current analysis module is configured to analyze the direction of fault current on the distribution network topology map based on a graph neural network to determine candidate fault regions; The interference cancellation module is configured to fuse data from the feeder terminal unit to verify the candidate fault area. The fault distance determination unit is configured to compare the voltage or current phasor changes of each node in the distribution network before and after the fault, and calculate the fault distance based on the impedance matrix. The traveling wave localization unit is configured to analyze the traveling wave generated when a fault occurs.

5. The self-healing control system based on multi-agent cooperation as described in claim 4, characterized in that, The fault location module includes: The information exchange module is configured to control multiple regional agents to exchange boundary measurement values ​​of the distribution networks under their jurisdiction. The step-by-step algorithm optimization unit is configured to perform iterative calculations based on boundary measurements using an alternating direction multiplier algorithm to solve for the cooperative localization result. The intelligent agent voting unit is configured to combine the cooperative localization results, the fault distance, and the traveling wave localization information to determine the fault location through a consensus algorithm. The agent voting compensation unit dynamically allocates the voting weights of the agents in the region based on the reliability of PMU data.

6. The self-healing control system based on multi-agent cooperation as described in claim 5, characterized in that, The self-healing control strategy generation module includes: The path search unit is configured to search for the optimal power supply restoration path based on the power distribution network topology map; The islanding unit is configured to generate a minimum islanding scheme containing a sequence of switching actions based on the optimal power restoration path. The resource allocation unit is configured to verify the minimum islanding scheme. If the distribution network is overloaded after adopting the minimum islanding scheme, a power balance scheme is generated through a smart negotiation algorithm. The issuing unit is configured to issue the minimum islanding scheme generated by the islanding unit, or to issue the minimum islanding scheme after power balancing adjustment.

7. The self-healing control system based on multi-agent cooperation as described in claim 6, characterized in that, The regional intelligent agent also includes: The adaptive control unit is configured to use a model predictive control algorithm to optimize the control action of the smart switch in the next few seconds.

8. The self-healing control system based on multi-agent cooperation as described in claim 7, characterized in that, The regional intelligent agent also includes a communication module; the communication module is configured to communicate with the local intelligent agent.

9. The self-healing control system based on multi-agent cooperation as described in claim 1, characterized in that, The system also includes a central decision-making agent configured to enable multiple regional agents to work collaboratively.

10. A self-healing control method based on multi-agent cooperation, characterized in that, The self-healing control system based on multi-agent cooperation as described in any one of claims 1-9, the method comprising: Set up regional intelligent agents and local intelligent agents in the power distribution network; The local intelligent agent is used to acquire real-time operating data of monitoring points in the power distribution network and to perform rapid fault detection on the real-time operating data. When the power distribution network fails, the regional agent integrates the fault analysis information of multiple local agents to determine the fault location information, and generates a self-healing control strategy based on the fault location information. When the distribution network fails, the local agent executes the self-healing control strategy issued by the regional agent on the distribution network.

Images