Parameters Identification of Photovoltaic Cell and Module Using LSHADE

Identifying circuit model parameters for photovoltaic cell and module is a challenging issue that often translates into an optimization problem. Recently, the most mainstream solution to such problems is based on metaheuristic optimization algorithms. Although there are many metaheuristic algorithms for the problem, the obtained parameters are often not very accurate and reliable. Therefore, a linear population reduction success-history based parameter adaptation for differential evolution (LSHADE) is applied to accurately and reliably identify the parameters of photovoltaic models. In LSHADE, the population size is continually decreased according to a linear function. The effectiveness of LSHADE is evaluated by identifying the parameters of the single diode model, the double diode model and the photovoltaic module model. The experimental results show that LSHADE outperforms other well-established parameters identification algorithms with respect to accuracy, stability, and rapidity.

I. Introduction

With the increasing consumption of non-renewable resources and the emergence of environmental pollution, people pay more and more attention to the development and utilization of renewable energy. Renewable energy refers to natural resources that can be maintained or increased by natural forces at a certain rate of growth. It includes solar energy, water energy, wind energy, biomass energy and so on. As a kind of natural energy, solar energy is abundant and pollution-free, showing its unique advantages. It has been internationally recognized as one of the most competitive energy sources in the future. Thereby, the photovoltaic (PV) systems, as the vital element for converting the solar energy into electrical energy through solar modules and other ancillary equipment, are extensively investigated by researchers in recent years [1]. In fact, the performance of PV systems is mainly affected by its model, thus an appropriate and accurate PV model is necessary and desired. For the mathematical models of PV system, although various models have been designed, the widely used models include the single diode model and the double diode model. Generally, the parameters of PV models have a great influence on its accuracy. Moreover, the performance evaluation and simulation of PV models, as well as the maximum power point tracking of PV systems require accurate models parameters. Hence, developing efficient algorithms to extract PV models parameters is still a significant task.