Electric energy meter self-inspection and fault prediction method based on edge calculation

Through edge computing, multi-scale empirical modal decomposition and adjacency graph construction are carried out on the power meter, combined with sparse gated residual graph attention layer and LSTM model, the problems of high data back-pass pressure and poor real-time performance in the power meter self-test and fault prediction are solved, and efficient fault detection and prediction are achieved.

CN120446858AActive Publication Date: 2025-08-08SHENZHEN FRIENDCOM TECH DEV +1
7 Cites 6 Cited by

Patent Information

Application Number
CN202510956491.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-07-11
Publication Date
2025-08-08
Estimated Expiration
2045-07-11

AI Technical Summary

Technical Problem

In low-voltage distribution networks, the self-test and fault prediction of power meters face the problems of high data return pressure, poor real-time performance and insufficient robustness. Especially in scenarios where decentralized deployment and unstable communications are not stable, traditional centralized fault detection methods are difficult to meet the needs.

Method used

Using an edge computing method, the self-test model package is sent to the edge nodes through the main station cloud platform, and a multi-scale empirical modal decomposition is performed to generate the local confidence vector, and an adjacency graph is constructed, combining the sparse gated residual graph attention layer and multi-layer LSTM model for fault prediction, realizing local self-test and cross-node fault prediction.

Benefits of technology

It reduces the data back-pass pressure, improves the real-time and system robustness of abnormal detection, realizes rapid judgment of the operating status of the power meter and multi-node coordinated fault prediction, and improves the accuracy and response speed of fault prediction.

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

Abstract

The invention belongs to the technical field of electric energy meters, and particularly relates to an electric energy meter self-inspection and fault prediction method based on edge calculation. The method comprises the steps that S1, a master station cloud platform issues a self-checking model package matched with the specification of an electric energy meter to each edge node installed in an electric energy meter concentrator; the self-checking model package comprises an encrypted handshake file and an inference engine; s2, the edge node completes communication connection with the master station cloud platform through the encrypted handshake file, then calls an inference engine, and confirms the multi-dimensional weight mask as a confidence coefficient vector and outputs the confidence coefficient vector; s3, after the master station cloud platform receives the confidence vectors of all the edge nodes, the output vectors are uploaded to the master station cloud platform, cross-node time sequence modeling is carried out by the master station cloud platform LSTM, and global fault prediction is generated; according to the invention, the real-time performance of anomaly detection and the robustness of the system are improved.
Need to check novelty before this filing date? Find Prior Art