Research On The Operation Strategy Of Electric Taxi Aggregator Considering The Uncertainty Of Multi-market Electricity Price And Individual Vehicle Behavior

Under the background of energy transformation,electric vehicles are an important demand-side resource,among which electric taxis have complex power consumption characteristics and are difficult to control,which is a difficult and blank point in demand response research.Under the electricity market environment,in order to realize the regulation and utilization of electric taxi load resources by the power grid,it is necessary to introduce a third-party main body electric taxi aggregator and study the operation mode and strategy of electric taxi aggregator participating in the electricity market.This paper studies the charging business mode and aggregator operation mode that are compatible with the load characteristics of electric taxis,focuses on various uncertainties in the decision-making process of electric taxi aggregators participating in the market and puts forward the model and method of aggregator’s operation strategy,which provides support for the bidding and scheduling decision-making of electric taxi aggregators in multiple markets.The main work of this paper is summarized as follows:The electricity market mode and charging business mode which are suitable for the load characteristics of electric taxis are comprehensively analyzed.In view of the decision-making problem of aggregators’ operation under the new charging business mode,this paper puts forward a data-driven optimization framework for the operation strategy of electric taxi aggregators,through which the integration of multi-source data,various uncertainties and various artificial intelligence methods can be realized,thus making up for the defect that the traditional model-based method can not adapt to the complex uncertainty optimization problem.In view of individual behavior randomness of electric taxi drivers and the coupling characteristics with the environment,the individual behavior simulation model of "electric taxi-operating environment" interaction is constructed,the event-driven complex interaction process simulation of "electric taxi-operating environment" is realized,and the individual behavior strategy of electric taxi based on reinforcement learning algorithm is proposed,which provides intelligent decision support for individual electric taxis in uncertain operating environment and provides simulation data support for the study of behavior-based load model and load model considering behavior uncertainty.In view of the lack of correlation characteristics of multi-market price fluctuations and "dimensionality disaster" in parameter estimation of multivariate models,a multi-market price uncertainty model is constructed,furthermore combining the conditional value-at-risk coupling theory,a bidding strategy model of electric taxi aggregators considering multimarket risk coupling under passive demand response scenario is constructed,and a multimarket bidding strategy solution method based on scenario generation and scenario reduction with consideration of correlation is proposed,which quantifies the influence of multi-market price correlation characteristic information on aggregators’ bidding decisions,improves the model credibility and computational efficiency of aggregator multi-market bidding strategy.In view of the time-series coupling characteristics of individual behaviors of electric taxis,the uncertainty model of individual charging decision based on time-series association rules mining method is proposed,on this basis,multi-stage bidding,scheduling strategy model and solution for electric taxi aggregators under active demand response scenario are proposed,the feasibility and economy of the optimization scheme and strategy of aggregators under active demand response scenario are discussed.