Heating, ventilating and air condition（HVAC）energy consumption on load side accounts for a large proportion of total energy consumption of the building. Adopting appropriate control scheme can effectively reduce energy consumption of energy consumption system. At present, the methods of energy saving regulation and control of HVAC on load side mostly focus on load forecasting control, which requires the combination of multiple model prediction results and high accuracy requirements. However, association rule analysis method can directly learn the law of historical operation data, and can more easily obtain more energy-saving regulation strategies. Energy consumption system of a shopping mall in Wuhan is taken as the research object, and operating data of the system load side in June are collected for research. After the data is cleaned and discretized, Apriori algorithm is used to calculate the association rules between power and device start-stop and frequency, and the energy-saving control scheme is presented based on this. Research results show that Apriori algorithm can learn the law of historical operation data of energy consumption system well. By analyzing the relationship between system power and adjustable parameters such as frequency, start and stop, the energy-saving control strategy can be obtained more quickly, conveniently and directly in actual use.