A novel global FPPT strategy for PV system under partial shading condition

Flexible power point tracking (FPPT) is a common and significate photovoltaic control problem under partial shading conditions. Based on CPP and MPP distribution analysis, both constant power point tracking and maximum power point tracking could be mathematically described by a single objective optimization. As one of the recent meta-heuristic algorithms, which avoid the limitations of conventional scanning methods, the carnivorous plant algorithm (CPA) solves non-convex optimization problems with fast speed, high accuracy, and simple calculation. However, it exhibits several drawbacks such as inappropriate parameters, over-searching, and feasible solution ignored issues. Therefore, a global FPPT strategy based on modified CPA (mCPA) is proposed, which includes a search objective integration, efficient non-convex search space skipping strategy, and adaptive tuning parameters according to sensitivity analysis. By statistical experiments compared with eight popular meta-heuristic algorithms, the mCPA can correctly converge to the global MPP under various scenarios with the highest accuracy of 99.0%. Meanwhile, the observations of dynamic experiments demonstrate the effectiveness of the proposed strategy in terms of fast time response (0.5-3 s) for global FPPT under various uniform and non-uniform insolation conditions. Based on statistical results (more than 2430 MATLAB simulations and hard-in-loop experiments) with a comprehensive comparison index, the CPA is selected and modified to serve as the core optimizer for the GFPPT optimization problem. Based on the features of MH and GFPP, a novel modification strategy is proposed to reduce searching time, including an integrated search objective, non-convex region skipping strategy based on equivalent or-der-preserving mapping, and adaptive tuning parameters. A novel and efficient mCPA-based GFPPT control algorithm is designed to automatically track the GMPP or CPP of a photovoltaic array under any PSC, which is also applicable to other MH-based methods.image