Fault electric arc detection method based on electric arc spectrum signals

Abstract

The invention relates to a fault electric arc detection method based on electric arc spectrum signals. The fault electric arc detection method comprises the following steps that a spectrum collector collects spectral signals which comprise the wavelength and the intensity of light according to set time and frequency, meanwhile, current signals are collected, the spectrum signals and the current signals are recorded, and normalization processing is conducted on the spectrum signals; then, wavelet transformation detail information is calculated to serve as characteristic values, meanwhile, the spectrum signals are integrated according to different wave bands, the wavelet transformation characteristic values and the spectrum intensity integral of the spectrum signals in different wave bands serve as the input of a BP neural network, and whether fault electric arcs exist or not is judged according to the output of the BP neural network. The fault electric arc detection method is wide in adaptability, high in safety and capable of directly reflecting the use conditions of an electric control switch and accurately judging the electric arcs.

Description

technical field [0001] The invention relates to an arc detection technology for actual circuit protection, in particular to a fault arc detection method based on arc spectrum signals. Background technique [0002] The rapid development of the electric power industry has brought a lot of convenience to people's daily life, but it has also brought frequent electrical fire accidents. my country's electrical fire method technology lags behind, and traditional detection devices cannot effectively detect arc faults. When arc faults occur, the instantaneous local high temperature will reach 2000 to 4000 degrees Celsius. If the arc fault is not cut off in time, it will cause serious consequences such as severe burning of electrical equipment, fire, and personal accidents. Therefore, the detection of fault arc is very important work. [0003] my country's research on arc detection technology started relatively late, and is currently mainly in the theoretical research stage (patent ...

AI Technical Summary

Problems solved by technology

[0003] my country's research on arc detection technology started relatively late, and is currently mainly in the theoretical research stage (patent application number: 201310376133.4, invention name: AC fault arc detection method based on wavelet transform and time-domain mixed features), most of them are detected by current signals, analyzed The current signal waveform is used to judge whether it is an arc fault, but the load conditions vary widely, and sometimes the difference between the current waveform when an arc fault occurs and the normal operation is very small (Zou Yunfeng, Wu Weilin, Li Zhiyong. Cluster analysis of low-voltage fault arcs based on self-organizing map neural network [J].Journal of Instrumentation, 2010, 31(3)571-576.), it is difficult to judge whether it is an arc fault

[0004] There are relatively mature arc fault detection products abroad, but the voltage environment at home and abroad is different, the arc characteristics will change, and the existing products cannot be directly applied. Therefore, it is very necessary to study the arc fault detection products suitable for China.

Images