A Novel Adaptive Flower Pollination Algorithm for Maximum Power Tracking of Photovoltaic Systems Under Dynamic Shading Conditions

Maximum power point (MPP) technique in photovoltaic (PV) systems implements a tracking controller, which is utilized to optimize the energy production under variable atmospheric conditions. The tracking process becomes more difficult due to appearance of many peaks owing to partial shading conditions. Although conventional and soft computing technologies are frequently used to solve MPP tracking issues, their performance is constrained by the fixed step size of conventional methods. However, once soft computing methods reach a certain MPP, they are constrained by a lack of randomness. The novel adaptive flower pollination algorithm (AFPA) optimization technique proposed in this work, proceeds with global and local searching in a single step, which is very crucial for the success of the MPP tracking with this method. The robustness of the approach is examined by conducting zero, weak, moderate, and strong shading patterns to a complete performance assessment via simulation, and that performance is compared with traditional flower pollination algorithm (FPA) and particle swarm optimization (PSO) techniques. This newly proposed method has the following advantages over the conventional FPA: a) risk of failure is zero; b) oscillation of power, voltage, and current across the load is minimized; c) produced energy is increased by 0.5 to 2.5% with respect to FPA; d) MPP is tracked smoothly and e) reduced MPP tracking time by average 45%. This advantage is especially noticeable in the dynamic variation of the shading patterns.