Photovoltaic (PV) systems are key contributors to renewable energy, yet their efficiency can be severely affected by partial shading conditions. These conditions introduce multiple peaks in the power-voltage (P-V) curve, complicating the task of conventional Maximum Power Point Tracking (MPPT) algorithms, which often fail to accurately locate the global maximum power point (GMPP). In recent years, metaheuristic approaches such as the Grey Wolf Optimization (GWO) algorithm have been applied to enhance MPPT performance. Despite their advantages, traditional Series-Parallel (SP) configurations are prone to local maxima, limiting their overall effectiveness in complex shading scenarios. This paper introduces an innovative control strategy that dynamically reconfigures the PV array between Series-Parallel (SP) and Total Cross-Tied (TCT) configurations based on real-time shading conditions. By adapting the PV array configuration and utilizing an optimized GWO algorithm, the proposed approach simplifies the P-V curve, facilitating more rapid and accurate convergence to the GMPP. The GWO algorithm was further optimized to enhance its search efficiency, reducing the risk of local maxima entrapment. Simulation results demonstrate that the proposed TCT-GWO configuration significantly improves both power output and convergence time in comparison to the SP-GWO configuration. In one case, the TCT-GWO approach increased power output by 37.3% over the SP-GWO configuration. These findings underscore the potential of dynamic PV array reconfiguration combined with metaheuristic optimization for improving the performance of PV systems under partial shading. This method offers a viable and effective solution for real-world PV applications, particularly in environments with frequent shading.