AC Arc Fault Detection Method of Internet of Things Terminal Based on Support Vector Machine

Abstract

Arc faults have become the main cause of electrical fires, causing heavy economic losses and casualties. In order to avoid the occurrence of fire, AC arc fault detection technology has been paid more and more attention. This paper presents an AC arc fault detection method of Internet of Things terminal based on support vector machine. Firstly, use Matlab to establish an AC arc fault simulation model, and obtain a large amount of current data during normal and AC arc faults through simulation; then, wavelet decomposition of the current data is carried out, and appropriate eigenvalues are extracted based on this calculation, and the percentage of the energy of each frequency signal to the total energy is caculated and used for the next training of the classifier; finally, the classifier model is established based on the support vector machine, and the python code is successfully used to implement the AC arc fault detection algorithm. The correctness of the algorithm is verified after the correct rate recognition test on the test set.