Intelligent fault monitoring method and system for power distribution network containing distributed energy

By combining graph generative adversarial networks and graph convolutional networks to extract node features, a Markov decision model is constructed, and an improved D4PG algorithm is used to train the agent. This solves the problem of multi-source data fusion and fault detection in dynamic environments in power distribution networks with distributed energy resources, and achieves fast and accurate fault identification and decision-making.

CN122085045APending Publication Date: 2026-05-26STATE GRID HENAN ELECTRIC POWER COMPANY ZHENGZHOU POWER SUPPLY CO
0 Cites 0 Cited by

Patent Information

Application Number
CN202511894077.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-12-16
Publication Date
2026-05-26

AI Technical Summary

Technical Problem

In distribution networks containing distributed energy resources, existing technologies struggle to effectively integrate multi-source heterogeneous data and capture the spatiotemporal coupling characteristics of the power grid. Furthermore, fault detection methods are ill-suited to complex dynamic environments, resulting in insufficient accuracy in fault identification and decision-making.

Method used

A framework combining graph generative adversarial networks and graph convolutional networks is used to extract node feature matrices, construct a Markov decision model, and train the agent using an improved D4PG algorithm. Experience replay is optimized through a multi-scale nearest neighbor algorithm and a dynamic feedback mechanism, and a generalized Huber loss function is designed to enhance the robustness and adaptability of fault detection.

Benefits of technology

It enables rapid and accurate fault detection in complex power distribution networks with active sources, improves the intelligence and reliability of fault decision-making, reduces false alarms and missed alarms, and enhances the timeliness and accuracy of detection.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses an intelligent fault monitoring method and system for a power distribution network containing distributed energy, and the method comprises the steps: firstly obtaining multi-source information data of the power distribution network containing the distributed energy, carrying out the feature extraction, and generating a node feature matrix; then, based on the node feature matrix, constructing a Markov decision model for fault detection of the source-containing power distribution network; taking a main body of the intelligent fault detection system as an intelligent agent, and training the intelligent agent by using an improved D4PG algorithm based on a Markov decision model; and after training is completed, exporting an optimal fault detection strategy of the source-containing power distribution network, and deploying the optimal fault detection strategy in a fault detection task of the source-containing power distribution network to perform intelligent fault monitoring of the power distribution network. Through a deep reinforcement learning framework, a complex source-containing power distribution network fault detection task is converted into an optimal decision problem, a direct, rapid and reliable logic mapping relation from a power grid fault state to an optimal diagnosis action is successfully established, and a more robust and adaptive fault detection method is provided.
Need to check novelty before this filing date? Find Prior Art