A comprehensive load series arc fault identification method

Abstract

The invention discloses a comprehensive load series connection arc fault identification method. The method includes the steps of obtaining a main circuit current feature information through a current sensor or obtaining branch circuit loading end voltage feature information signals through voltage sensors, conducting A / D sampling at a fixed sampling frequency, then, filtering noise out of the sampled signals through a filtering algorithm, analyzing the main circuit current feature information signal and branch circuit loading end voltage and arc fault feature information and comparing the main circuit current feature signal and the branch circuit loading end voltage and arc fault feature information with preset threshold values, judging whether arc faults may occur to the voltages of the loading ends of the branches and the current of the main circuit or not, determining whether arc faults occur to a circuit by combining the arc fault judgment result of the branches and the main circuit with the arc fault judgment result of arc faults with relatively weak features after comprehensive parallel connection is achieved, and determining the specific branch circuit according to the voltages of the branch circuits. According to the method, the loading series connection arc faults with obvious fault features under the single loading or the comprehensive loading can be detected, and the load series connection arc faults with relatively weak fault features under the comprehensive loading can be detected.

Description

technical field [0001] The invention relates to the field of electric circuit fault detection, in particular to a comprehensive load series arc fault identification method. Background technique [0002] In the fire accidents over the years, the proportion of electrical fires has always been high, and arc faults are one of the important causes of electrical fires. The existing arc fault detection and identification methods are single, and the generalization ability of arc fault identification for distribution line nonlinear loads or inductive loads is not strong, and it is easy to cause misjudgment of faults in practical applications; The arc fault of the load branch also has the problem of insufficient reliability of missed judgment. Therefore, this patent proposes an identification technology for series arcs under comprehensive loads, aiming to eliminate the problem of missed or misjudgment existing in existing arc fault detection methods. [0003] At present, the researc...

AI Technical Summary

Problems solved by technology

However, the normal operating current of nonlinear loads has a "zero break" phenomenon similar to arc fault currents, and the starting current of inductive loads has the characteristics of a large rate of rise. These characteristics are very similar to the sudden rise after the arc current crosses zero. The existing arc fault current identification method based on the extraction of post-zero mutation features does not have a strong generalization ability for arc fault identification of distribution line nonlinear loads or inductive loads, and it is easy to cause misjudgment of faults in practical applications; in addition, using arc faults alone The current detection method still has some problems of missed judgment. For example, if the current transformer is placed in the main circuit and the loads of the two branches are connected in parallel, if an arc fault occurs in one of the loads, the fault current of some loads will be submerged in the main circuit. In the arc fault detection, it may be filtered out as a noise interference signal, so that the arc fault current detection set in the main circuit cannot effectively identify the arc fault of the branch circuit

Therefore, none of the existing arc fault detection methods can effectively deal with loads with relatively weak fault characteristics after comprehensive parallel connection. Solving this kind of situation then fire hazard still exists

Images