To contribute to flexible loads participating in the demand response in high renewable energy permeability distribution network, a clustering method based on electric vehicles (EVs) aggregation for virtual power plant (VPP) is proposed in this paper. The proposed clustering method is based on charging characteristics parameters from multiple load profiles of electric vehicles. For the VPP bidding strategy, aggregation model based on clustering method, economic scheduling model considering EVs aggregation, inelastic loads, wind power resource, photovoltaic resource, distributed generators and energy storage devices. In this paper, a linearized VPP model for multi-resources participating in day-ahead scheduling under active distribution network is proposed. The presented model is a linear mixed-integer programming and solved the optimization solver. The proposed aggregation method can increase the economics for VPP and promote the consumption of renewable energy of distribution network by applying the model to the IEEE 33-bus feeder.