The large-scale integration of distributed generator and the extensive application of electric vehicles(EVs) not only optimize the energy structure

but also bring new challenges to siting and sizing of distribution network.In this paper

the wind speed

light intensity and demand load samples with correlations are obtained through combining the Latin hypercube sampling

Cholesky decomposition and Nataf transformation.The Monte Carlo simulation is used to obtain the charging load of the electric vehicles with randomness.Combined with the wind-charge sample considering the uncertain factors

the typical operation scenarios are obtained by K-means clustering method.Based on the chance-constrained programming

the programming model which is solved by the constriction factor particle swarm optimization algorithm is established.The IEEE 33-bus distribution network is applied to analyze the influence of correlation and different charging modes of EVs on the planning scheme.The results show that

the grid load can be reduced if the EVs chose the conventional charging mode

and considering the correlation between wind

photovoltaic and load can enhance the economic efficiency of the planning scheme and make it more instructive.