Improving gain of real time PI controller using particle swarm optimization in active power filter

Abstract

This paper incorporates the PSO based PI control for systematic switching function of APF topology. The proportional and integral (PI) gain tuning of active power filter (APF) using the particle swam optimization (PSO) method for reactive power compensation and harmonics mitigation. A conventional PI controller required more computation time and has less accuracy. The harmonic load currents are extracted using an instantaneous real and reactive power scheme. The performance indices including total harmonic distortion, reactive power, power factor and capacitor voltage regulation using PSO trained PI controller is compared with convention PI controller. PSO has quick convergence, minimal adjusting parameters and fast execution to solve non-linear problems. The traditional PI controller is substituted by an online PSO trained PI controller for the objectives of enhancing the dc voltage tracking in APF under non-linear load conditions. The proposed work is developed in sim-power system tool box which is a software package in Matlab/ Simulink.