Parameters Extraction of Photovoltaic Models Using Triple-Phase Teaching-Learning-Based Optimization

Abstract

Parameters extraction of photovoltaic (PV) models is urgently desired for the simulation, control, and evaluation of PV systems. To accurately and reliably extract the parameters of different PV models, a triple-phase teaching-learning-based optimization (TPTLBO) is proposed in this paper. The novelty of TPTLBO lies in: i) teaching-learning-based optimization introduces a buffer phase and adopts a centroid strategy to update the position of intermediate learners, which further strengthens the exploration and exploitation; ii) the learners can select different phases and employ different learning strategies based on their knowledge level; iii) a dynamic control parameter replaces the original random parameter rand to enhance the search ability of algorithm. The parameters extraction performance of TPTLBO is verified through the single diode model, the double diode model, and three PV models. Experimental results demonstrate that TPTLBO achieves better performance in terms of accuracy and reliability compared to state-of-the-art algorithms.

Keywords

• Photovoltaic models
• parameter extraction
• teaching-learning-based optimization
• triple-phase