Research On Electric Vehicle Charging Control Strategy Based On ADRS Multi-energy Optimization

Based on the research background of electric vehicle(EV)high efficiency and energy-saving charging control,supported by Harbin Science and technology innovation talents research special fund project,An EV multi energy charging optimization control technology based on automatic demand response strategy(ADRS)is developed.Taking the lowest charging cost of electric vehicles as the optimization objective,the clustering analysis,parameter prediction and other related algorithms are developed,This paper studies an optimal control strategy of EV charging based on ADRS theory to meet the urgent demand of energy saving and consumption reduction control for large-scale EV charging under the current carbon peaking and carbon neutralization demand.The main research contents are as follows:The control mechanism of EV charging optimization is analyzed by combining ADRS with power supply circuit modeling.The multi energy complementary optimization power supply system of conventional power system and photovoltaic power generation is established to meet the high power and high efficiency power supply demand of EV,and to realize the optimization of EV charging power and cost performance under the condition of multi energy power supply.The research can realize the multi energy coordinated planning between the utility system and photovoltaic power generation,and provide the basis of mechanism analysis and control algorithm for high efficiency EV charging demand.In order to realize the optimal allocation of charging station resources,a clustering multi node learning Gaussian process(CMNL-GP)method is proposed to realize the optimal control of EV charging.The proposed CMNL-GP uses the shared and node specific parts of all nodes to fuse the historical data from different nodes.Under similar conditions,the proposed CMNL-GP model predicts multiple charging point nodes,and gives the topology combination of charging points that can meet different EV loads.The effectiveness of the proposed CMNL-GP algorithm is verified by analyzing the measured data of EV charging stations within 24 hours.In order to analyze the negative effects of parallel charging of multiple charging piles on the power grid,a charging load model of electric vehicles based on Monte Carlo idea is established,and the charging behavior is clustered and defined by the branch and bound theory,and the negative effects of parallel combination of multiple charging piles are modeled and analyzed.The calculation results show that the proposed orderly charging strategy can fill the load trough of power grid,reduce the peak valley difference of power grid,and effectively reduce the impact of EV charging on the load of connected power grid.This research combined with CMNL-GP can be used as the basis of algorithm analysis of ADRS optimal control.The dynamic price vector generation model based on K-means clustering method is used to generate the price vector in the case of no prediction algorithm,and realize the development of automatic demand response algorithm under the condition of real-time change of real-time price and voltage amplitude,which is used to minimize the power cost purchased from the grid,and ensure the self consumption rate of photovoltaic power generation and the satisfaction degree of charging demand.Simulation results based on dynamic price vector generation model and dynamic feasible energy demand area model show that the proposed ADRS strategy has good control effect in reducing EV charging cost.