Abstract: It is one of the effective ways to accelerate the development of pumped storage by integrating and transforming the existing cascade hydropower stations and adding pumped storage units to form cascade hybrid pumped storage power stations. Compared with conventional cascade hydropower stations, cascade hybrid pumped storage power stations have new pumping operation conditions. It has a more complex cascade hydraulic connection than the traditional pumped storage power station. The flexible operation mode requires a more refined scheduling modeling method. This paper proposes a short-term peak shaving optimization model of a cascade hybrid pumped storage power station to explore its typical peak shaving operation mode. The model aims to minimize the maximum peak-valley difference of the remaining load of the power grid and takes the unit as the minimum dispatching unit. Differentiated modeling is adopted for the pumping-generating operating conditions of different types of units. Regarding model solving, the original nonlinear model is transformed into a mixed integer linear programming (MILP) model by a linear transformation method and then solved by the CPLEX mathematical tool in the JAVA environment. Taking a cascade hydropower station in southwest China as an example, the analysis results show that the peak-valley difference of residual load of cascade hybrid pumped storage power station is reduced by 4.6 % compared with that of conventional cascade hydropower stations.