Online selective harmonic minimization for cascaded Half-bridge multilevel inverter using artificial neural network

Abstract

Selective harmonic elimination (SHE) is a widely used method to minimize the amount of total harmonic distortion (THD) in multilevel inverters. The main challenge associated with SHE method is to solve multiple variable high order nonlinear equations to obtain desired switching angles. To overcome this problem, a new online approach is provided in this paper which does not require solving complex nonlinear equations. In this method, artificial neural networks (ANNs) are used to determine the switching angles to minimize the THD amount of the ac side voltage of an 11-level cascaded Half-bridge (CHB) multilevel inverter. However, unlike offline methods, ANN is updated online, without using any data gained from solving SHE equations. Finally, Simulation results are illustrated and compared with other offline methods.