A method and apparatus for constructing a macro- and micro-parameter correlation model of an electric arc

Abstract

This invention discloses a method and apparatus for constructing a macro- and micro-parameter correlation model of an electric arc. The method includes: Step S1, establishing a macroscopic characteristic model of the fault arc, obtaining the time-domain and frequency-domain calculation results of the arc current, and analyzing its effectiveness; Step S2, establishing a micro-multivariate model of the fault arc, obtaining the calculation results of the multivariate micro-model of the arc current, and analyzing its effective values; Step S3, based on the calculation results of the macroscopic characteristic model and the micro-multivariate model of the fault arc, constructing a correlation model of the macro- and micro-parameters of the arc using Bayesian networks and causal forests. This invention utilizes Bayesian network methods to analyze the correlation strength of macro- and micro-parameters of the arc and employs causal forest methods to obtain clear correlation relationships. This solves the problem of errors easily generated during processing due to the large amount of data obtained from the model, and provides a more organized and efficient comparison of data at both the micro and macro levels, thereby drawing accurate conclusions.

Images