Abstract

With the development of society and the progress of science and technology, the market of photovoltaic power generation is gradually turning to grid-connected power generation system. However, the output power of photovoltaic power generation is vulnerable to the interference of the external environment, and its volatility and intermittent shortcomings will impact the main power grid. Therefore, it is of great significance for the prediction research of photovoltaic power generation. In this paper, the photovoltaic of smart microgrid project of shanghai electric power university is taken as the research object, a back propagation neural network (BP) model based on particle swarm optimization (PSO) and momentum method is proposed, and the weight and bias of BP neural network are optimized by PSO algorithm and momentum method. In the early stage, the data were firstly normalized, and then the important features were selected through the gray correlation degree analysis. Finally, the photovoltaic power prediction experiment was conducted. The results show that the algorithm model in this paper has better prediction performance for photovoltaic power generation.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call