Performance Evaluation of Solar MPPT Technique using Fast and Improved PSO (FIPSO) Under Partial Shading Conditions

Partial shading is a menace to the PV power generation systems which is practically unavoidable. Non-uniform shading can be due to clouds, tree branches, buildings or other nearby items. The non linear output characteristics... [ view full abstract ]

Partial shading is a menace to the PV power generation systems which is practically unavoidable. Non-uniform shading can be due to clouds, tree branches, buildings or other nearby items. The non linear output characteristics of a PV source varies with temperature and solar insolation. Whenever partial shading condition occurs, the conventional hill climbing methods of MPPT prove to be inefficient. During partial shading the power-voltage (P-V) curve exhibits multiple peaks. Conventional MPPT methods get stuck on one of the local maxima and fail to attain the global maximum power point (GMPP). Particle swarm optimisation (PSO) has proved to be a very accurate and powerful technique to find GMPP. In this paper we improve the conventional PSO by reducing swarm size, as well as, the number of iterations to achieve MPP at a fast pace with remarkable accuracy, thus making our system more efficient with less computation requirements. Simulation and performance evaluation of the proposed technique under different partial shading conditions prove its advantages, such as flexibility, reliability, system-independence and high accuracy in tracking the GMPP under non uniform insolations.

The first figure (Particle Initialisation (4 particles)) shows the initialisation of the the swarm particles, which are very close to the local maxima occurring in the P-V characteristics. The second figure (Convergence in 5 Iterations) shows that in just 5 iterations, using the FIPSO algorithm,  the particles have managed to converge at the global maximum.