Abstract: As the main energy-consuming entities of intelligent buildings, heating, ventilation, and air conditioning（HVAC） systems have great significance to fulfil the flexible trade-off between their energy consumption cost and user comfort. However, when the above two factors are considered in parallel, the coupling terms of the system optimization model and the difficulty in solving increase. The learning-based control strategy is convenient in model construction, but the energy-saving effect is average. According to the above challenges, a bi-level optimal control strategy for HVAC system with collaborative consideration of air quality and thermal comfort is proposed. Firstly, the coupling relationship between temperature and air quality in each area is accurately portrayed based on the thermal dynamic model of the RC equivalent circuit and the internal physical structure of the building. Secondly, the operation strategy of the HVAC system is optimized with the objectives of the minimum system energy cost and maximum user comfort. A bi-level optimization approach is developed to solve this complex coupling problem, which optimizes the supply air volume at the upper level and the ventilation rate at the lower level, respectively. The error is corrected by using the rolling optimization method. Finally, the results of the proposed strategy with different comfort coefficients are analyzed in the summer cooling scenario and compared with other control strategies. The results show that the method can balance the economy and user comfort.