Otimização de ângulos de chaveamento para Conversores Multinível utilizando Enxame de Partículas

Abstract

This article presents an approach using Particle Swarm Optimization (PSO) to optimize the switching angles of a Multilevel Converter (MLC) in order to mitigate selected harmonics composing the converter's output voltage while maintaining a prescribed modulation index value. A special representation using real numbers is employed for the converter's switching angles to prevent unfeasible solutions from being considered during the process. A population heuristic guided by the Euclidean distance between candidates is applied to maximize the coverage of the initial solution space. The choices of parameter values and metrics applied in the PSO are justified through simulations in Matlab software. The analysis of the converter's output voltage waveform demonstrates the effectiveness of the proposed solution.