Adaptive search-space reduction and partial shading detection for high-efficiency MPPT in photovoltaic systems

This paper proposes an integrated maximum power point tracking (MPPT) framework for photovoltaic systems operating under uniform and partial shading conditions. The framework incorporates a partial shading detection technique (PSDT) that simultaneously identifies the operating condition and adaptively reduces the search space by eliminating regions that cannot contain the global maximum power point (GMPP) and avoiding redundant sampling, thereby accelerating the subsequent MPPT process. Based on the identified condition, an enhanced divide-and-conquer algorithm with ternary ratio (EDCA-TR), designed to avoid unnecessary inner-point evaluations, is deployed under uniform shading conditions, while a modified JAYA (M-JAYA) algorithm is employed under partial shading conditions. Unlike conventional iteration-based metaheuristic approaches, the proposed M-JAYA employs particle-based updates that immediately incorporate newly acquired search information, together with an adaptive search-space confinement mechanism based on iteration count and particle power information. Experimental validation under seven shading patterns demonstrates that the proposed framework achieves an average tracking time of 1.86 s and an average tracking efficiency of 95.27%, outperforming the Hybrid, improved coot optimization algorithm (ICOA), and improved water cycle optimization algorithm (IWCO) benchmark methods. Moreover, the proposed framework achieves an average tracking energy loss of 3.45 W·s, representing a reduction of more than 50% compared with the benchmark methods. Repeated-run simulation analysis further demonstrates generally consistent performance despite the stochastic nature of the M-JAYA algorithm. Overall, the proposed framework improves convergence speed, tracking stability, and energy extraction by combining condition-dependent MPPT selection with adaptive search-space reduction and particle-based optimization.