This Extraction of the maximum electrical power from a solar photovoltaic (PV) system under numerous weather conditions is required to reduce its payback time period, per unit energy price, and to compensate for the high initial price of the solar PV system. This could only be achieved by continuously operating the solar PV system at its maximum power point (MPP) under several weather conditions. Unlike under uniform weather conditions (UWC), identification of the real MPP (Global MPP) under partial shading condition (PSC) in a reasonable time is a challenging task due to the formation of multiple local MPP in the power-voltage (P-V) characteristic curve of a solar PV array. The natureinspired MPP tracking algorithms have been proved suitable for global MPP tracking (MPPT) under PSC. In this research paper, a renowned nature-inspired flower pollination algorithm (FPA) is deeply reviewed, modified, and integrated with the random walk filter to improve its performance in terms of tracking speed, and efficiency. A comparison of the proposed 'Adaptive Flower Pollination Algorithm (AFPA)' and conventional FPA algorithm has been made under zero, weak, and strong PSCs for a 4S solar PV array. The proposed algorithm has produced remarkable results in tracking speed, and efficiency, for the global MPP (GMPP) tracking under different PSCs. The simulation is performed in MATLAB/Simulink software.