Abstract: Considering the adverse impact of the disorderly access of flexible loads such as electric vehicles （EVs） and air conditioning loads on the power grid, an elastic optimal scheduling method for load in supply region is proposed, which takes into account the situation analysis of EV power supply resources. Firstly, the probability prediction of the charging demand of EVs is carried out. Through quantitative analysis of the characteristic indicators of EV loads, a situation awareness model for regional EV power supply resources is proposed, and the situation assessment of power supply resources is performed through  training of integrated learning algorithms. Then, according to the situational awareness result, this paper proposes an optimal scheduling strategy that flexibly controlling the charging demand of EVs. It takes the charging demand of EVs and the reduction amount of air conditioning load as control amounts, establishes a multi-objective scheduling optimization model with elastic constraints, and improves multi-objective particle swarm optimization. The algorithm is solved to obtain the optimal scheduling plan. Finally, the analysis of regional examples verifies that the proposed optimal scheduling method can achieve peak regulation of the power load, coordinately resolve the conflicts between the demand for flexible loads and idle resources, orderly dispatch   EVs charging  and air-conditioning   load to maximize the utilization of power supply resources.

Key words: electric vehicle charging requirements, situation awareness of power supply resources, integrated learning, flexible optimal scheduling