MPPT Perturbation Optimization of Photovoltaic Power Systems Based on Solar Irradiance Data Classification

• Published in: IEEE transactions on sustainable energy 2019-04, Vol.10 (2), p.514-521
• Main Authors: Yan, Ke, Du, Yang, Ren, Zixiao
• Format: Article
• Language: English
• Subjects: classification; Clouds; Deserts; Irradiance; machine learning; Maximum power point trackers; Maximum power point tracking (MPPT); Maximum power tracking; Optimization; Perturbation; Perturbation methods; Photovoltaics; PV power system; Sea measurements; Solar power; support vector machine (SVM); Support vector machines; Testing; Training
• ISSN: 1949-3029; 1949-3037
• DOI: 10.1109/TSTE.2018.2834415

The tracking accuracy and speed are two main issues for the fixed step perturb-and-observe maximum power point tracking (MPPT) method. This study proposes a novel solution to balance the tradeoff between performance and cost of the MPPT method. The perturbation step size is determined off-line for a specific location based on the local irradiance data. The support vector machine is employed to automatically classify the desert or coastal locations using historical irradiance data. The perturbation step size is optimized for better system performance without increasing the control complexity. Simulations and experiments have been carried out to verify the effectiveness and superiority of the proposed method over existing approaches. The experimental results show a 5.8% energy generation increment by selecting optimal step sizes for different irradiance data types.