A simulation of multiple peaks maximum power point tracking for photovoltaic system

Recent decades, renewable energy applications have shown significant increase due to the awareness from people regarding the negative impacts of fossil fuel burning to the environment. The availability of renewable energy source also has contributed to the increase in renewable energy applications. Among the renewable energy sources available, photovoltaic (PV) is considered to be one of the best. This is mainly because it provides clean energy silently. However, there are some limitations to PV applications. Considering that it has a high initial cost to set up and it has very low conversion efficiency, it is vital to ensure that a PV system can operate at Maximum Power Point (MPP). Maximum Power Point Tracking (MPPT) controller such as Perturb and Observation (P&O), Incremental Conductance (Inc. Cond), and Hill Climbing (HC) are widely used due to simple implementation and shows a good performance in tracking MPP when solar irradiance is uniform. However, when partial shading occurs on the PV array, tracking to MPP becomes complicated as multiple peaks exist on the Power-Voltage (P-V) characteristic curve. Conventional methods cannot distinguish between local peaks and global peak, thus failed to track the true MPP. Several methods based on stochastic algorithm and artificial intelligence has been developed to track MPP under partial shading conditions. This project focuses on the performance of MPPT controller to extract maximum power from PV system under partial shading condition. The selected MPPT algorithms that have been implemented in the PV system includes Fuzzy Logic Controller and Particle Swarm Optimization. Results show that both the simulated MPPT controllers are capable of tracking the maximum power up to 90% of its efficiency.