Taking the substation building as the research object, DeST software is used to simulate the annual building load of the case building, and the load results are connected to the TRNSYS software. At the same time, the G ENOPT software is used to optimize the solar inclination and azimuth, and the system simulation model of solar photovoltaic-air source heat pump is established and the data analysis is carried out. The results show that the peak cooling load of the substation building occurred on July 24th, The numerical value is the 49.28kW, The annual cumulative cooling load of the building is 20165.59 kWh; Based on Hooke Jeeves algorithm, the optimal inclination of photovoltaic panels is 44.18 and the optimal azimuth is 11.69; Typical days, The highest peak power of the air source heat pump appears at 1 a.m,Its peak power is 7.47 kW, The COP peak was 3.67; Within the typical week, Solar photovoltaic efficiency shows stable periodic changes, Peak aking around 12:00 noon, Its peak efficiency is 18.14%; There is a significant linear distribution relationship between the solar photovoltaic power generation capacity and the solar radiation, Its linear correlation coefficient was 0.85. Compared with the single air source heat pump system, the composite system saves 18.33% energy and has good application effect.