Fault detection algorithm of inverter based on singular value decomposition nonlinear filtering

Abstract

Power electronic devices, power semiconductor devices and control circuits most were vulnerable to failure. In this paper, took the inverter as example, discussed its fault type, and designed a new fault detection algorithm based on singular value decomposition (SVD) nonlinear filtering. Simulation result shown that this algorithm could detect fault conveniently, and using extended Kalman Filter (EKF) based on singular value decomposition (SVD), stability of filter could be evaluated, at the same time, calculated amount was not increase, and filter precision was not decrease. This algorithm could meet the need of fault detection.