The energy consumption of air conditioning system accounts for a large part of building energy, and the optimal control strategy plays a vital role in system energy saving. Taking a cooling system of a company in Shanghai as an example, the verification model is established on the TRNSYS-MATLAB co-simulation platform based on the energy data collected by intelligent sensors. A time-division optimization strategy for integrated load forecasting control and indoor temperature fuzzy control is proposed. Finally, the total energy consumption of the system under this strategy is compared by the other two control strategies, namely, the water supply temperature (SWT) control strategy and the all-day room temperature fuzzy (RTF) strategy. Result shows that under the load forecasting integrated (LFI) strategy, the cooling system can save 16.44 % energy without sacrificing thermal comfort.