Selection of Step Size based on Weather Type Classification for Maximum Power Point Tracking in Photovoltaic System

A tradeoff between dynamic and steady-state performance can be solved by using variable step-size maximum power point tracking methods. However, the choice of the optimal N value in the previous research is rarely investigated. Thus, in this work, the selection of the value of N is designed by using weather type classification. The K-means algorithm is used to choose the typical weathers of difference classification from whole year data in the desert area. This selection method of N value can improve efficiency without increasing the control complexity, and the effectiveness of the proposed selection of N value has been verified through experimentation.