CN-120999889-B - Substation operation and maintenance method, device, equipment and medium based on multiple intelligent agents

CN120999889BCN 120999889 BCN120999889 BCN 120999889BCN-120999889-B

Abstract

The embodiment of the disclosure discloses a substation operation and maintenance method, device, equipment and medium based on multiple agents. The method comprises the steps of responding to the fact that the body quantity of a transformer substation is higher than a target body quantity, obtaining a real-time operation data set and an agent coordination diagram, conducting diagram division on the agent coordination diagram to obtain at least one agent group set, executing generation steps for each agent group set, generating group decision information for each agent group, displaying the group decision information set on a decision page, generating an agent decision diagram, executing agent screening of the coordination agent set to obtain a coordination agent subset and agent decision information, generating operation and maintenance information, and controlling a main agent to instruct each coordination agent according to the operation and maintenance information so as to conduct equipment operation and maintenance treatment on the target transformer substation. According to the method and the device, under the condition of reducing the calculation resources of the intelligent agent, accurate generation of the corresponding operation and maintenance information of the transformer substation can be realized, and normal operation of the transformer substation is guaranteed.

Inventors

ZHANG YEFENG
ZHANG LUYUN
JIAO FENG
WANG ZENGHAI
XIONG RUI
YANG CHENJIE
ZOU YING
CHEN JING
CHEN XIAODONG
JU BIN
LI XIAO
JIANG YUANKUN
LIU YUAN
HOU RUI

Assignees

国网信息通信产业集团有限公司
国网信息通信产业集团有限公司北京分公司
北京电力经济技术研究院有限公司
华北电力大学
国网思极紫光(青岛)云数科技有限公司

Dates

Publication Date: 20260508
Application Date: 20250729

Claims (7)

1. A multi-agent based substation operation and maintenance method, comprising: In response to determining that the body quantity of a transformer station corresponding to a target transformer station is higher than a target body quantity, acquiring a real-time operation data set and an agent coordination diagram corresponding to the target transformer station, wherein the agent coordination diagram corresponds to a main agent and a coordination agent set; Responding to an agent processing operation of an operation and maintenance object in an agent coordination diagram corresponding to a decision page, and adjusting each agent node and an agent side in the agent coordination diagram according to the agent processing operation to obtain an adjusted agent coordination diagram, wherein the decision page supports clicking operation of each agent node in the agent coordination diagram, and pops up a popup window corresponding to the agent node after clicking the agent node, wherein the popup window displays at least one of associated equipment, associated agent nodes, environmental states, execution action sets and confirmation information of whether to participate in operation and maintenance; Determining the adjusted agent coordination diagram as the agent coordination diagram, wherein each agent node in the agent coordination diagram is a node for confirming whether to participate in operation and maintenance; carrying out graph division on the agent coordination graph by utilizing the main agent to obtain at least one agent group set, wherein each agent group set is divided based on target division characteristics, and each agent in the agent group carries out coordination decision; for each agent group set, the following generating steps are performed: for each intelligent agent group, generating group decision information corresponding to the intelligent agent group according to the real-time operation data group corresponding to the intelligent agent group; displaying the group decision information set on a decision page corresponding to the main agent; generating an agent decision graph corresponding to the agent group set according to the candidate decision information determined on the decision page, wherein the agent decision graph supports various visual operations; According to at least one agent decision diagram, performing agent screening of the coordination agent set to obtain agent decision information corresponding to the coordination agent subset and the remaining coordination agent set; generating operation and maintenance information corresponding to the target transformer substation according to the agent decision information and the coordination agent subset; And controlling the main intelligent agent to instruct each coordination intelligent agent according to the operation and maintenance information so as to perform equipment operation and maintenance treatment on the target transformer substation.
2. The method of claim 1, wherein said mapping the agent coordination graph with the master agent to obtain at least one agent group set comprises: Acquiring operation and maintenance direction information corresponding to the target transformer substation input in the decision page; performing feature disassembly on the operation and maintenance direction information to obtain at least one target division feature and at least one corresponding feature division direction; for each target partitioning feature, the following partitioning steps are performed: determining a target feature division direction corresponding to the target division feature; Carrying out graph division on the agent coordination graph according to the target division characteristics by using the main agent to obtain an initial agent sub-graph set; determining initial decision adjustability corresponding to each agent in the initial agent sub-graph set in the target feature division direction by using the main agent; Displaying the obtained initial agent sub-graph set on a page jumped by the decision page corresponding graph dividing control, so as to confirm and adjust each dividing graph under the target dividing characteristic and the adjustability corresponding to each initial decision; acquiring an agent sub-graph set adjusted for the initial agent sub-graph set and a corresponding decision-making adjustability group set; And determining the intelligent agent group set corresponding to the target division characteristic from the intelligent agent group set corresponding to the intelligent agent sub-graph set.
3. The method of claim 2, wherein the generating the group decision information corresponding to the agent group according to the real-time operation data group corresponding to the agent group comprises: setting a reward and punishment function corresponding to the intelligent agent group according to the characteristic dividing direction corresponding to the intelligent agent group; And generating group decision information corresponding to the intelligent agent group by utilizing the reward and punishment function according to the decision adjustability group and the real-time operation data group corresponding to the intelligent agent group.
4. The method of claim 1, wherein the method further comprises: In response to determining that the body quantity of the transformer substation corresponding to the target transformer substation is not higher than the target body quantity, determining whether the resources meet the requirements of each agent corresponding to the agent coordination diagram or not according to the resources corresponding to the server cluster corresponding to the target transformer substation, and performing coordination output to obtain operation and maintenance information; In response to determining that the target importance condition is not met, screening at least one agent with importance degree meeting the target importance condition from the agent coordination diagram, wherein resources required by the coordination output corresponding to the at least one agent are smaller than resources corresponding to the server cluster; And generating and executing the operation and maintenance information of the at least one agent by using the average field theory and reinforcement learning according to the real-time operation data set, and taking the operation and maintenance information as the operation and maintenance information corresponding to the target transformer substation.
5. A multi-agent based substation operation and maintenance device, comprising: The acquisition unit is configured to acquire a real-time operation data set and an agent coordination diagram corresponding to a target transformer substation in response to determining that the transformer substation volume corresponding to the target transformer substation is higher than the target volume, wherein the agent coordination diagram corresponds to a main agent and a coordination agent set; The device further comprises an agent processing operation performed by responding to the operation and maintenance object on an agent coordination diagram corresponding to a decision page, and according to the agent processing operation, each agent node and an agent side in the agent coordination diagram are adjusted to obtain an adjusted agent coordination diagram, wherein the decision page supports clicking operation of each agent node in the agent coordination diagram, after clicking the agent node, a popup window corresponding to the agent node is popped up, and at least one of the following items is displayed, namely association equipment, association agent nodes, environment states, and confirmation information of whether an action set participates in operation or not, wherein each agent node in the agent coordination diagram is a node for confirming whether the operation and maintenance participate or not after adjustment; The division unit is configured to divide the agent coordination diagram by utilizing the main agent to obtain at least one agent group set, wherein each agent group set is divided based on target division characteristics, and each agent in the agent group performs coordination decision; The execution unit is configured to execute the following generation steps of generating group decision information corresponding to the intelligent agent group according to the real-time operation data group corresponding to the intelligent agent group for each intelligent agent group, displaying the group decision information set on a decision page corresponding to the main intelligent agent, and generating an intelligent agent decision graph corresponding to the intelligent agent group according to candidate decision information determined on the decision page, wherein the intelligent agent decision graph supports various visual operations; the screening unit is configured to execute the agent screening of the coordination agent set according to at least one agent decision graph to obtain agent decision information corresponding to the coordination agent subset and the rest coordination agent sets; the generation unit is configured to generate operation and maintenance information corresponding to the target transformer substation according to the agent decision information and the coordination agent subset; And the control unit is configured to control the main intelligent agent to instruct each coordination intelligent agent according to the operation and maintenance information so as to perform equipment operation and maintenance treatment on the target transformer substation.
6. An electronic device, comprising: One or more processors; a storage device having one or more programs stored thereon, When executed by the one or more processors, causes the one or more processors to implement the method of any of claims 1-4.
7. A computer readable medium having stored thereon a computer program, wherein the program when executed by a processor implements the method of any of claims 1-4.

Description

Substation operation and maintenance method, device, equipment and medium based on multiple intelligent agents Technical Field The embodiment of the disclosure relates to the technical field of computers, in particular to a substation operation and maintenance method, device, equipment and medium based on multiple intelligent agents. Background At present, with the continuous development of electric power, the transformer substation is maintained in time, so that the timely supply of electric power can be effectively ensured, and the occurrence of electric power fault condition is avoided. For maintenance of the transformer substation, the method generally comprises the steps of firstly, based on reinforcement learning, indicating the corresponding intelligent agents of the transformer substation to carry out overall decision output, and obtaining operation and maintenance information corresponding to the transformer substation. Then, the operation and maintenance information corresponding maintenance operation is executed. However, when the above manner is adopted, there are often the following technical problems: Under the condition that the transformer substation corresponds to a large number, the number of the intelligent agents corresponding to the transformer substation is large, the decision output is integrally carried out, the intelligent agents correspond to a large combined action space, the linkage action control grows exponentially, a large amount of server resources are large, and the conditions of memory and computing resource exhaustion occur. The above information disclosed in this background section is only for enhancement of understanding of the background of the inventive concept and, therefore, may contain information that does not form the prior art that is already known to those of ordinary skill in the art in this country. Disclosure of Invention The disclosure is in part intended to introduce concepts in a simplified form that are further described below in the detailed description. The disclosure is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter. Some embodiments of the present disclosure propose multi-agent based substation operation and maintenance methods, apparatuses, devices, and mediums to address one or more of the technical problems mentioned in the background section above. In a first aspect, some embodiments of the present disclosure provide a substation operation and maintenance method based on multiple agents, including, in response to determining that a target substation corresponds to a substation volume higher than the target volume, acquiring a real-time operation data set and an agent coordination map corresponding to the target substation, where the agent coordination map corresponds to a main agent and a coordinated agent set, using the main agent to map the agent coordination map to obtain at least one agent group set, each agent group set being partitioned based on a target partitioning feature, performing a collaborative decision for each agent in the agent group set, generating group decision information corresponding to the agent group for each agent group according to the real-time operation data set corresponding to the agent group, displaying the group decision information set on a decision page corresponding to the main agent, generating an agent decision map corresponding to the agent group decision map according to candidate decision information determined on the decision page, performing a collaborative decision for each agent according to the at least one agent group, and performing a collaborative decision for each agent group according to the target coordination map, and performing a collaborative decision information for each agent group, and performing a collaborative decision for each agent group according to the target coordination map. In a second aspect, some embodiments of the present disclosure provide a multi-agent-based substation operation and maintenance device, comprising an acquisition unit configured to acquire, in response to determining that a target substation corresponding substation amount is higher than a target amount, a real-time operation data set and an agent coordination map corresponding to the target substation, the agent coordination map corresponding to a main agent, a partitioning unit configured to map the agent coordination map using the main agent to obtain at least one agent group set, each agent group set being partitioned based on a target partitioning feature, each agent in the agent group making a collaborative decision, an execution unit configured to perform, for each agent group, a generation step of generating, for each agent group, group decision information corresponding to the agent group based on the real-time operation data set corresponding to the agent group, displaying the grou