Electric vehicle and distributed wind generation (DWG) accessing to grid are hotspot of current research. Firstly, the stochastic characteristics of DWG and electric vehicle to grid (V2G) are illustrated, and the method of distribution network load modeling including electric vehicle charge/discharge power is proposed based on nonparametric kernel density estimation, which could better overcome the error brought by traditional normal distribution in load modeling. Secondly, the wind penetration limit calculation model is constructed based on chance constrained programming (CCP), and the particle swarm optimization (PSO) algorithm embedded with stochastic simulation is proposed to solve the calculation model. Finally, the influence of V2G on the distribution of distribution network load is stated on a tested system; what is more, the influence of electric vehicle penetration on the wind penetration limit is analyzed from multiple perspectives.