In the context of the large-scale connection of distributed photovoltaic（PV） to the power system, in order to make use of the complementary characteristics of the distributed pumped storage power plant with other energy source output and maximize the output of combined PV-pumped storage system to meet the scheduling requirements while reducing the abandoned PV power, the output models for distributed photovoltaic and distributed pumped storage units that are more in line with the actual operation conditions are established. On this basis, a joint scheduling model of PV and pumped storage is developed and a deterministic transformation approach is used to solve the scheduling model with random variable prediction errors in the objective function. The Euclidean distance is introduced as a measure of the matching degree of the scheduling curve. Finally, example simulation results show that the proposed strategy can make the distributed PV output better meet the scheduling demand with the cooperation of pumped storage, which promotes the system's consumption of distributed PV power.