Online Evolutionary Maximum Power Point Tracking for Photovoltaic Grid-connected System

The utilization of clean energy is increasing because of energy shortage. Meanwhile, the advancement of photovoltaic grid-connected technology has emerged as one widespread approach both domestically and internationally as a crucial component of microgrids. Thus, this study proposes an online evolutionary strategy for photovoltaic maximum power point tracking to fully utilize the solar energy in grid-connected photovoltaic and realize adaptive adjustment of maximum power point tracking control facing environmental dynamics. In this study, the conventional regulation is upgraded into an online evolutionary approach, which deploys incremental adjustment and deep learning. A deep neural network is proposed to continuously modify the weight matrix, which adaptively tunes tracking parameters. In addition, a two-stage single-phase grid-connected photovoltaic inverter is designed with environmental dynamics. The proposed strategy of online evolutionary maximum power point tracking is demonstrated on photovoltaic panels under different solar conditions. The experimental results present that the developed evolutionary approach can overcome the shortcomings of traditional maximum power point tracking, which accelerates the response speed and reduces the power fluctuation under external disturbances.