Method for fusing and diagnosing fault information of circuit of electric meter on basis of SOM (self-organized mapping) and D-S (Dempster-Shafer) theories

Abstract

A method for fusing and diagnosing fault information of a circuit of an electric meter on the basis of SOM (self-organized mapping) and D-S (Dempster-Shafer) theories includes firstly, creating a fault mode set for the faulty circuit according to circuit analysis of the faulty circuit of the electric energy meter and fault modes specified by GJB299C; secondly, selecting to-be-observed fault signal points corresponding to fault modes in the set according to the fault mode set created in the first step and using the to-be-observed fault signal points as test points for functions and states of the circuit; thirdly, preprocessing fault signals acquired at the fault signal points selected in the second step; fourthly, fusing fault information by the aid of the SOM theory, outputting fault conclusions, selecting 70% of data for training and selecting 30% of the data for testing; and fifthly, fusing the fault conclusions by the aid of the D-S theory and making a decision for faults. By the aid of the method, confidence degree of a fault diagnostic result is further increased, integral uncertainty caused by errors is reduced, accuracy of fault diagnosis is greatly improved, and the method is an extremely important means in the field of information fusion.

Description

(1) Technical field: [0001] The present invention provides a kind of electric energy meter circuit fault information fusion and fault circuit diagnosis method, especially relate to a kind of electric energy meter circuit fault information fusion based on self-organizing feature map network (that is based on SOM) and based on D-S theory (D-S theory is developed by Dempster in It was first proposed in 1967 and further developed by his student Shafer in 1976 as an imprecise reasoning theory, also known as evidence theory or D-S evidence theory) for the diagnosis of electric energy meter circuits. It belongs to the field of circuit fault information fusion and diagnosis. (2) Background technology: [0002] In the past ten years, as a typical application of system fault simulation technology in the field of circuit design—the comprehensive analysis technology of circuit performance, reliability and testability based on fault simulation has achieved rapid development. In this tech...

AI Technical Summary

Problems solved by technology

[0005] In this patent, the feature layer fusion classifies the input data through SOM and calculates the trust degree of the classification results. However, when the amount of evidence increases or the classification results conflict, convincing diagnostic results cannot be obtained only by relying on SOM

Images