Power dispatch data network fault locating method and system based on sparse autoencoder

By constructing a hierarchical mesh topology model with a sparse autoencoder and an SAE fault location system, the problems of low fault location accuracy and high false alarm rate in the power dispatch data network were solved, achieving efficient fault equipment location and visualization, and improving the reliability and operation and maintenance efficiency of the power dispatch data network.

CN122113571APending Publication Date: 2026-05-29STATE GRID SHANGHAI MUNICIPAL ELECTRIC POWER CO
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN Β· China
Patent Type
Applications(China)
Current Assignee / Owner
STATE GRID SHANGHAI MUNICIPAL ELECTRIC POWER CO
Filing Date
2026-01-12
Publication Date
2026-05-29

AI Technical Summary

Technical Problem

Existing fault location methods for power dispatch data networks suffer from poor topology adaptability, unreasonable dataset construction, limited feature representation, and lack of visualization, resulting in low location accuracy, high false alarm rate, and poor engineering practicality.

Method used

A hierarchical mesh topology model based on a sparse autoencoder is constructed to generate a real-world fault training dataset. This dataset is then used to train an SAE fault location model, which is then visualized to achieve accurate location of faulty equipment.

Benefits of technology

It improved the accuracy of fault location, reduced the false alarm rate, shortened the fault handling time, and enhanced the engineering practicality and visualization capabilities of the model.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122113571A_ABST
    Figure CN122113571A_ABST
Patent Text Reader

Abstract

The application relates to a power dispatching data network fault positioning method and system based on a sparse self-encoder, which comprises the following steps: constructing a hierarchical mesh topology model of a power dispatching data network; simulating a single-device fault scene based on the hierarchical mesh topology model to generate a fault training data set; training an SAE fault positioning model with sparse constraints by using the fault training data set; collecting real-time device state data and real-time link state data of the power dispatching data network to be measured, and preprocessing the real-time device state data and the real-time link state data to obtain a real-time fault data set; inputting the real-time fault data set into the trained SAE fault positioning model to obtain a fault device identifier and a confidence degree, and obtaining the position of a fault point in the power dispatching data network. Compared with the prior art, the application solves the technical problems of poor adaptability and high false alarm rate of a traditional fault positioning method, and provides reliable technical support for rapid disposal of a power dispatching data network fault.
Need to check novelty before this filing date? Find Prior Art