Parameter estimation of four diode solar photovoltaic cell using hybrid algorithm

Abstract

ABSTRACT The effective and precise parameter estimation of the solar photovoltaic (PV) cell is extremely crucial for precise evaluation and control of PV systems. In the recent years, many meta-heuristic and conventional algorithms have been proposed for the parameter estimation of PV cells. However, it is still a huge challenge for researchers. In this study, a new hybrid algorithm Hybrid Particle Swarm Optimization and Dingo Optimizer (HPSODOX) based on two widely used meta-heuristic algorithms i.e., Particle Swarm Optimization (PSO) and Dingo Optimizer (DOX) is developed for parameter estimation of four diode model of solar PV cell under different operating conditions. Two main components of any hybrid algorithm are exploration and exploitation, and HPSODOX has the ability to make fine balance between these components. The hybrid algorithm is benchmarked on ten functions, out of which seven are unimodal, and three are multimodal functions. Further, results achieved with developed hybrid algorithm for parameter estimation of four-diode solar PV cell model are compared with those of three meta-heuristic algorithms i.e., PSO, GWO, DOX and two hybrid algorithms i.e., GWOCS, PSOGWO. The result of four diode model is compared with single, double and triple diode model. For the four diode model the computational time of the proposed hybrid algorithm (HPSODOX 2.09) is better than the rest of the compared algorithms (PSO 4.75, GWO 4.73, DOX 4.43, GWOCS 2.25, and PSOGWO 2.23). The comparative analysis show that HPSODOX performed significantly better than other algorithms with minimum error and best optimal solution. After the extraction of parameters, the non-parametric test is performed i.e., Friedman ranking test. The results so obtained clearly indicate that the proposed hybrid algorithm is better than the rest of the compared algorithms.