Electric Vehicles Charging Management Research And Simulation Software Development Based On Multi-Agent

Many studies have shown that disordered charging behaviors of electric vehicles(EVs) will bring many negative effects to the power systems, such as power grid load fluctuation increase, lower load rate etc., which will directly affect the stability and economic operation of the power systems, and will lead to higher requirements of the power systems including generation, transmission and distribution capacity of power.This paper presents a multi-objective integrated charging management strategy based on Multi-Agent model to improve the impact of EVs on the grid and achieve charging management of EVs. A simulation software platform is designed based on JADE.The main work of this paper can be divided into the following four parts:(1) Analyze the current situation EVs development and its influence. First, future car’s development is EVs-oriented; second, the causes and the adverse effects of EVs on the power systems are analyzed; finally, the importance of controlling the charging power to charge EVs is emphasized according to the orderly arrangement from the generation, transmission, and distribution of solving the perspective of the adverse effects on the grid.(2) Analyze Agent model and Multi-Agent Model of charging network. The notion of Agent is introduced including its theories, communication, coordination strategies,and the principles of JADE management platform based on Multi-Agent; then, building a Multi-Agent model of charging network, and layered Agents are discussed in detail including the software models of each part, responsibilities and mechanism of mutual coordination principle. The feasibility and advantages of Multi-Agent model for charging network will be presented.(3) Analyze the algorithm designed for charging management strategy. Considering the interaction between the user and the power systems, a new multiple targets algorithm design are selected and multi-objective and PSO algorithm are added to weight overall satisfaction approach to establish an objective function in order to achieve a comprehensive evaluation of the charging process.(4) Comprehensive analysis of the overall design and the results from simulation software. Combining with JADE management platform, a Multi-Agent simulationsoftware based on Agent software model is developed for EVs charging network, and choose simulate data which is close to reality as possible according to charging conditions as raw data of software. A Multi-Agent instance of charging network is established by simulation software. The validity, reliability and accuracy of the Multi-Agent charging network and management strategy are verified according to the simulation results.