How to orderly charge electric vehicles with randomness is an urgent problem to be solved in“green travel”period today. Firstly, a mathematical model of electric vehicles based on advanced data mining technology is established. To grasp the load characteristics of the electric vehicle, the fuzzy C-means clustering algorithm is used to analyze the driving behavior of the owner and the initial battery state of the incoming electric vehicle taking the historical data of electric vehicle travel in a community in Beijing as an example. Based on the idea of Markov chain, a model consisting of several typical electric vehicle load mode transition probability matrices is established to describe the stochastic dynamic process of electric vehicles. Multiple population GA is applied to solve the orderly charging problem of electric vehicles.Taking the community charging station as an example and comparing with other methods, the effectiveness of the model and method is verified.