A Novel Adaptive Control Strategy for Transient Performance Improvement of DC/DC Converter in Distributed Power Generation Systems

This paper presents an adaptive predictive correction control method for DC-DC converter based on sequential neural network. In this method, NARX network is used to predict the output voltage in the transient process of the converter, and the reference value of the controller is compensated and corrected, so as to realize the adaptive control of the converter. Simulation results verify the feasibility of the proposed method. Compared with traditional PI control, the proposed method can effectively suppress voltage fluctuation in the transient process and significantly shorten the adjustment time of the converter, which is expected to be applied in high-performance regulated voltage power supply and distributed power generation.