Abstract: For the consumption problem of new energy in northwest China,the paper integrates the wind power,photovoltaic power,concentrating solar power plant and electric energy storage devices to form a virtual power plant（VPP）,and proposes a multi-time scale stochastic optimization scheduling strategy of new energy virtual power plant based on robust stochastic optimization theory. First,mathematical description of the wind power photovoltaic power,concentrating solar power plant and electric energy storage devices is established. Based on that,the VPP multi-time scale optimization scheduling model is established. In the day-ahead scheduling layer,the paper takes the VPP operation benefit maximization as the goal,and establishes the day-ahead optimal scheduling model based on day-ahead forecasting output power of the wind power and photovoltaic power. In the hour-ahead scheduling layer,operation costs minimum is taken as the goal,and the day-ahead scheduling correction model is established by forecasting the current output of the wind power and photovoltaic power. Simultaneously,a VPP random scheduling model is established by application of measuring the impact of output uncertainty of wind power and photovoltaic power on system operation. The results show that the model can improve operating profit and new energy consumption capacity.