Research On Electric Vehicle Charging Behavior Based On Multi-Agent Simulation Technique

As energy resource crisis and environmental pollution problems are improving, electric vehicle with its advantages of high energy utilization rate has been regarded as one of an important solution to global energy crisis and protect global environment, so developing electric vehicles has become pressing needs of present.However, large-scale electric taxi accessing will cause great influence to power grid operation and planning, especially the disordered and dispersed electric vehicle charging behavior will pose a threat to the safe and stable operation of power grid. So it is necessary to research the optimized control of electric vehicle charging behavior and solve the problems brought by accessing to power grid.The demonstration operation of electric taxi is significant to guide the development of electric vehicle. This paper focuses on electric taxi, establishing the Q-Learning model of electric taxi agent from three aspects of state space, behavior decision space and behavior revenue function. We presented a multi-agent-based simulation system of electric taxi charging behavior on JADE platform, describing the information interaction and mutual influence among agents such as electric taxi, transportation network and charging station, and through simulation, we analyze and compare the electric taxi charging behavior decision-making under two kinds of behavior revenue model. The simulation results show that this system can reflect the characteristic of electric taxi charging load, which provides references for research the planning and operation of electric vehicle.Secondly, through the simulation system, we can obtain the distribution of electric taxi charging load under different charging price strategy. The example analysis shows that this system can reflect the response of electric taxi charging behavior to the changing charging price, which provides references for reasonable policy of electric vehicle charging price.Finally, under the problems of present siting of charging stations, the comprehensive evaluation model of different scheme in the new siting of charging stations is established based on Analytic Hierarchy Process, which considering operation fee, charging distance and charging waiting time. This model provides new practical method for optimal siting of electric vehicle charging station.