Optimal reconfiguration strategy based on modified Runge Kutta optimizer to mitigate partial shading condition in photovoltaic systems

In this article, a new variant of a recent optimization algorithm namely Runge Kutta optimizer (RUN) is proposed to solve the partial shading condition in photovoltaic systems and global optimization. The RUN's improvement is mainly to mitigate the lack of the solutions quality, the imbalance between the exploitation and exploration phases, and premature convergence of the RUN algorithm. The mRUN, in which the standard RUN is incorporated with a promising strategy namely Orthogonal learning (OL) strategy to address out the original RUN's drawbacks. To estimate the proposed algorithm's efficiency, two experimental series were performed. In the first experiment, the mRUN was compared with other state-of-art metaheuristics on IEEE CEC’2020 test suite. Moreover, the quantitative and qualitative approved the robustness of the suggested algorithm. As a second experimental series, the proposed approach is investigated on an application of renewable energy-based system which is reconfiguration of partially shaded photovoltaic (PV) array. The main target of this process is to enhance the generated power from the PV array. Two shade patterns are investigated on 9 × 9 array and the obtained results via the modified RUN are compared to total cross tied (TCT) and SudoKu and other metaheuristic-based arrangements of Aquila optimizer (AO), Harris hawks optimizer (HHO), and Runge Kutta optimizer (RUN). The generated power from the PV array, arranged via the proposed approach, is enhanced by 28.41% and 1.015% in the first shade pattern compared to the TCT and conventional RUN configurations, respectively. In the second shade pattern, the proposed mRUN succeeded in improving the extracted power by 40.32% and 0.29% compared to TCT and RUN configurations, respectively. The modified RUN performed well in both studied cases outperforming the others in achieving the most enhanced power from the array.