Improved Particle Swarm Optimization MPPT for Standalone PV System Under Varying Environmental Conditions

A photovoltaic (PV) system is highly sensitive to dynamic changes in environmental conditions. Improving the maximum power point tracking (MPPT) algorithm is one of the most cost-effective ways to enhance its performance. Currently, the most widely used MPPT algorithm is Particle Swarm Optimization (PSO). However, as with many other artificial intelligence (AI) algorithms, PSO tends to become stagnant as it converges on the optimal solution and is therefore incapable of adapting to dynamic environmental changes. This paper presents an improved PSO MPPT called iPSO consisting of an intelligent mechanism to detect and adapt to dynamic environmental changes. In addition, to accelerate MPPT convergence, the inertia weight ($\omega$) of iPSO is decreased exponentially with increasing iterations. Furthermore, a deterministic initialization method (DIM) is employed to improve the probability of locating the global maximum power point (MPP). The MATLAB/Simulink platform is utilised to evaluate the performance of the proposed iPSO under various environmental conditions. Its convergence speed and tracking efficiency are evaluated and compared to those of standard PSO and PSO with a reinitialization mechanism called PSO-reinit. Overall, the results revealed that iPSO is 2.2 s faster and 5.7% more efficient than its closest competitor, PSO-reinit.