PSO, an effective tool for harmonics elimination and optimization in multi-level inverters

Abstract

So far, several methods have been presented to eliminate and minimize the low order voltage harmonics at multi-level inverter's output. Harmonics optimization aims to eliminate the low order harmonics, if it is possible, or at least to minimize them. Thus, harmonics optimization can be considered as a generalized form of harmonics elimination. In this way, this paper applies an advanced variation of particle swarm optimization method to 7 to 17-level inverters. Results are compared with continuous genetic algorithm, as a well-known intelligent method, and one of the most successful numerical methods, namely sequential quadratic programming. Results confirm that PSO completely outperforms both CGA and SQP for all cases.