Parameter extraction of photovoltaic modules using Newton Raphson and simulated annealing techniques

Accurate simulation of photovoltaic (PV) modules is strongly dependent on parameters that are not given in manufacturers' data sheet. Number of parameters depend on type equivalent circuit models: Rs-, Rp-, and 2-diode models with four, five and seven unknown parameters respectively. Availability of different parameter extraction techniques has been presented and broadly classified as analytical, iterative and evolutionary techniques each of which has its own advantages and disadvantages. For Rp model with five unknowns, analytical method may not give accurate results unless with some assumptions. To solve this problem Newton Raphson is supposed to give better result at the cost of difficulty in close approximation of initial conditions and increased iterations. Recent trends show that evolutionary algorithms give improved result but common drawbacks of increased computation time and iterations. This paper focused on parameter extraction for five parameter (Rp-) model using Newton Raphson and Simulated Annealing techniques followed by verification with experimental data extracted from manufacturers' datasheet. The I-V characteristic curves, absolute relative error plots and maximum power point matching capabilities revealed superior performance of Simulated Annealing over Newton Raphson technique.