Logo image
Research on intelligent energy management method of multifunctional fusion electric vehicle charging station based on machine learning
Journal article   Peer reviewed

Research on intelligent energy management method of multifunctional fusion electric vehicle charging station based on machine learning

Tao Shi, Fang Zhao, Hangyu Zhou and Caijuan Qi
Electric power systems research, Vol.229, p.110037
04/2024

Abstract

Deep learning Electric vehicle Energy storage Forecasting decision Photovoltaic power generation Reinforcement learning
•Reserch highlight 1:A typical physical architecture of the multifunctional charging station with photovoltaic power generation and battery energy storage was designed. Then considering the market price signal, charging load, PV power fluctuation and other uncertain factors, an energy management model was established to maximize the overall operating benefits.•Reserch highlight 2:An intelligent energy management architecture based on machine learning was proposed in order to improve the intelligence level of charging stations to achieve multi-functional integration and optimal operation of charging stations in the most convenient and economical way.•Reserch highlight 3:Power prediction algorithm based deep learning was implemented to meet the predictive requirement of charging station operation scenarios under uncertain conditions.•Reserch highlight 4:.Optimization decision making algorithm based deep reinforcement learning was implemented to meet the optimization decision-making requirement of energy management for multifunctional charging stations under uncertain conditions. The machine-learning based approach to energy management of multifunctional charging stations that meets the needs in the context of "carbon neutrality". The method takes multifunctional charging station as the research object, considers the uncertainty elements such as market price signal, charging load, and the change of photovoltaic power generation, and aims at maximizing the comprehensive operation efficiency of charging station in the decision cycle, realizes the prediction of power generation and load power by using deep learning method, and obtains the optimal energy management strategy by reinforcement learning, and realizes intelligent management and operation of multifunctional charging station. A typical case study is presented and simulation analysis is carried out for typical cases to verify the correctness and effectiveness of the method. It is shown that the method can effectively cope with the uncertain elements in the operation of multifunctional charging stations, reduce the hardware configuration and manual operation and maintenance costs of charging stations while ensuring safety constraints, and effectively improve the energy utilization efficiency and economy of the whole charging station.
url
https://doi.org/10.1016/j.epsr.2023.110037View
Published (Version of record) Open

Metrics

1 Record Views

Details

Logo image