Fault detection and line selection method of series arc fault in three-phase motor with frequency converter circuit

Abstract Series arc fault (SAF) can cause electrical fires. Aiming at the problem that in the industrial field, there is no effective protection method for the SAF, a fault detection and line selection method for the SAF was proposed based on stochastic resonance (SR) theory. First, the SAF experiments were conducted in the three-phase motor with frequency converter load (TPMFCL) circuit. Second, the SR was used to highlight the SAF features in the A-phase current signal at the front end of the frequency converter. The wavelet coefficients of each layer of the highlighting signal were obtained by using wavelet transform, then the energy of the obtained coefficients was calculated, and utilized as SAF features, and an identification model was developed by using support vector machine to identify the SAF. The optimal parameters of the above methods were determined by particle swarm optimization algorithm. Finally, the test was carried out and the SAF detection and line selection accuracy in the TPMFCL circuit can exceed 95% and 90%, respectively. It can provide a new idea for arc fault circuit interrupters with fault line selection function.