Optimal allocation for charging piles in multi-areas considering charging load forecasting based on Markov chain

Considering the complexity and diversity of user travel habits, and satisfying the charging demands, charging piles in different areas should be reasonable planned. Firstly, Markov chain is used to describe changes in battery state of charge with three decision-making behavior including driving, charging, neither charging nor driving of electric vehicle owner's trip in the whole day. According to real-time charging behavior in the process, the fast and slow charging demand of different vehicle types are predicted. Then considering mobility characteristics of electric vehicles and the number of different types of electric vehicles in different time periods in some area, the total load demand would be predicted. The optimization model aims to minimize investment and operating and maintenance costs for the charging piles, which takes balanced electric vehicle movement characteristics inequality constraints into account, and is realized by particle swarm optimization algorithm. The effectiveness of the proposed method is verified by the simulation on 33-bus system.