An array detection method for partial discharge based on sparse decomposition theory and RD-MUSIC

Abstract

Partial discharge (PD) detection is important to many scholars around the world. As an applicable method of PD detection, the ultrasonic array detection combines the ultrasonic array sensor and the array signal processing technique to detect PD. The PD ultrasonic array detection is affected by field noise and traditional PD direction algorithms generally face various problems such as low precision and a complex computational process. Considering this, we proposed a PD ultrasonic array detection method based on the de-noising method in spare decomposition theory and the reduced dimension multiple signal classification (RD-MUSIC) algorithm. Firstly, the matching pursuit (MP) algorithm in spare decomposition theory is improved to conduct simulations of PD signals; secondly, direction-finding is simulated for PD signals by using the RDMUSIC algorithm. Finally, the experimental data are subjected to de-noising and direction-finding by using spare decomposition theory and the RD-MUSIC algorithm after conducting an experimental study under laboratory conditions. Moreover, a comparison is made between the location results after, and before, de-noising to verify that the method is practicable.