Abstract: Firstly,based on k-means clustering and Copula theory,the conditional forecast error distribution model of the PV output power under different weather types is established,and the model is used to modify the forecast value of PV output power in the time series simulation algorithm.Secondly,taking the tracking planned output power of the PV power station as the control objective and taking the maximum annual average income in the life cycle of the PV power station as the optimization goal,considering the whole life cycle cost of energy storage,the penalty cost of deviating from the schedule and the profit of the PV power station,the particle swarm optimization and the time series simulation method are used to search the best configuration of the energy storage.Finally,taking the energy storage configuration scheme for a grid-connected PV power station with a rated capacity of 100 MW in Anhui Province as an example,the proposed method is validated.The simulation results show that the economy of the PV power station is improved and the power and the capacity of the energy storage are reduced when the modification is considered.