Combination of quadratic ranking selection real-coded genetic algorithm with the Hooke-Jeeves optimization method for solar photovoltaic parameter estimation

Abstract

Estimating accurate parameters for Photovoltaic cell (PVC) models plays a very important role in evaluation, simulation, control and maximum power point tracking (MPPT) of solar PV systems. Indeed, parameter extraction for solar cells and modules from experimental data is a challenging problem. These parameters must be evaluated from a set of nonlinear equations that cannot be solved analytically. Therefore, an optimization approach is essential to estimate these parameters by minimizing the difference between the calculated current from the PV model and the measured current. In this paper, a real-coded genetic algorithm (GA) based on quadratic ranking selection combined with the Hooke-Jeeves (HJ) method is proposed for this parameter extraction. It is well known that GA has proven good capacities in global search while on the other hand, the HJ has high-performance capacities in local search. As a hybrid technique, the HGAHJ has the feature of combining the capabilities of global and local searching. The proposed algorithm is developed and then applied for the parameter estimation for the single and double diode solar cell models. In addition, a comparative study with the GA and particle swarm optimization (PSO) algorithms is presented to demonstrate the effectiveness of the proposed approach using measured data of two commercial solar panels, which are Kyocera KC200GT and RTC France. Results reveal the superior capabilities of the suggested HGAHJ algorithm as a global optimization technique.