Optimized Energy-Storage Method Based on Deep-Learning Adaptive-Dynamic Programming

Abstract

Renewable energy has the characteristics of fluctuation and intermittence due to environmental conditions; the fluctuation of output power affects the operation of the power grid. To combat these fluctuations, a control strategy for the battery energy-storage system using a deep-learning adaptive-dynamic algorithm is proposed in this work. First, power-fluctuation rate feedback control is used to suppress the power fluctuation from the renewable energy source. Second, by introducing an adaptive-dynamic algorithm based on a deep belief network, the charging and discharging power of the battery energy-storage system with secondary regulation is achieved. Finally, the validity of the methods is verified by a series of examples using available data for wind and photovoltaic sources. It was found the proposed control strategy is highly effective for suppression of unwanted fluctuations.