Due to the corrosion and aging caused by the special oceanic environment, the characteristic of coastal photovoltaic (PV) system significantly drift after years of operation. In this study, the maximum power point tracking (MPPT) problem for coastal PV system is addressed and a novel MPPT methodology based on deep neural network (DNN) integrated with the corrosion evaluation index (CE-index) and dynamic training-sample (DTS) mechanism is developed. To be specific, the detailed effect of corrosion and aging for the PV modules installed in coastal areas is comprehensively analysed, and a composite indicator for evaluating the PV parameter drift, namely CE-index, is proposed. Then, a novel DNN-based offline MPPT methodology for the large-scale coastal PV system is developed, in which the DTS mechanism is also introduced for overcoming the effect caused by PV module corrosion and aging phenomenon. Finally, the optimal length of DTS for different degrees of CE-index is comprehensively verified by case studies. Experimental result shows that the developed DNN-based MPPT methodology can accurately forecast the maximum power point (MPP) voltage for large-scale coastal PV-system with robust performance, and cooperation of the developed DTS-mechanism and CE-index corrosion evaluation strategy can also effectively overcome the disturbance caused by the harsh oceanic environment.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.