| Summary: | Due to the influence of mutative environmental conditions, the photovoltaic (PV) array of a PV system receives with non-uniform irradiation and temperature, which leads to the power-voltage (P-V) output characteristic appearing multi-peak and the current-voltage (I-V) output characteristic emerging multi-steps. With the assistance of various optimization algorithms, maximum power point tracking (MPPT) technologies have become an effective method to improve the conversion efficiency of the PV system under different weather conditions. However, the recognition ability of these algorithms for global peak are still not guaranteed under uneven irradiation and temperature, which have attributed to absence randomness for these algorithms after reaching the maximum power point (MPP) region. Therefore, a modified flower pollination algorithm (MFPA) is proposed in this paper for MPPT. In MFPA, switching between dual-mode optimization is affected by both switch probability and population fitness values, and therefore overcomes the defects that the flower pollination algorithm (FPA) falls easily into the local maximum and slowly convergences in the later period. The performance of MFPA for MPPT is verified by comparing with the perturb & observe method and FPA. Simulation experiment results show that the proposed algorithm can rapidly and accurately track the MPP under various environmental conditions, especially the performance being superior under the condition of strong irradiation and partial shading.
|
|---|
