MPPT ALGORITHM BASED ON PARTICLE SWARM OPTIMIZATION WITH NATURAL SELECTION

Abstract

The output of photovoltaic array under non-uniform light is nonlinear with multi peaks. Since the traditional MPPT algorithm based on particle swarm optimization (PSO) is likely to fall into premature convergence and may work slower in the later iterations, the MPPT method based on improved PSO with natural selection was proposed for the first time in the paper. New algorithm worked by eliminating and replacing the particles with lower fitness during the iterative process and hence the swarm could track better and faster for the maximum point, with less chances of premature convergence. During the MATLAB modeling and simulation, the improved MPPT algorithm could perform better than the traditional PSO method.