A direct-drive PV pump system based on phase change heat storage is proposed to address the current building energy consumption problem in heating/cooling. The system can solve the problems of unstable solar energy, high battery cost, and short battery service life. Taking a 45m
2 office in Shanghai as the research object, the annual dynamic simulation model of the system is built through TSNSYS. The linear weighted evaluation function method is used for multi-objective optimization analysis of system performance and cost. The results show that the selection scheme with a PV area of 26 m2, a phase change volume of 0.14 m3, and a valley electricity storage ratio of 1 has the best-combined performance and cost of the system.