In order to solve the problems in generation planning and scheduling of power grid brought by the uncertainty and volatility of large-scale intermittent energy, for the strong volatility of wind as well as photovoltaic power and grid load, an economical scheduling strategy based on integration of wind, photovoltaic, water and thermal power has been proposed. The generalized load is defined and the flexible scheduling of pumped storage is used to translate the output space of intermittent energy to smooth the generalized load curves and fluctuation in power. The objective functions that the output of the integration made up of wind, PV and water is maximum and the fluctuation of generalized load is minimum are introduced into the process of modeling. Through comparison and analysis of simulation results, the proposed model can not only increase the efficiency of intermittent energy and realize the function of peak load shifting, but also decrease the startup and shutdown of thermal units to reduce the thermal economical costs.