In order to improve the integration of photovoltaic power generation in power systems

this paper proposes a carbon trading based scheduling model of hybrid energy storage system consisting of battery energy storage systems and pumped storage units in power systems with large-scale photovoltaic power generation. Based on the concept of low-carbon economy

the ladder-type carbon trading mechanism is introduced to the economic dispatch of power systems. In this model

the output scenarios of the photovoltaic power generation system are effectively clustered by the improved K-means clustering algorithm based on the maximum and minimum distance criteria

to reduce the number of scenarios under the premise of ensuring the output distribution characteristics of photovoltaic power generation system. With the goal of the lowest system operation cost

the operation economy and low carbon of the system

the battery energy storage is taken as power storage to smooth output fluctuations of photovoltaic power plants

and the pumped storage is taken as energy storage to participate in peak-shaving balance after the access of photovoltaic power generation system. Taking the improved IEEE-RTS96 system as an example

the proposed model is simulated and analyzed. The result of the example verifies the rationality and effectiveness of the proposed model.