On joint control of heating, ventilation, and air conditioning and natural ventilation in a meeting room for energy saving

In this paper, we consider the optimal joint control of heating, ventilation, and air conditioning systems and natural ventilation during the start-up period of these systems in a meeting room. The joint control policy could reduce the total energy consumption in the building, but the optimal policy could be complex and difficult to implement in practice. In order to address this dilemma, we make the following major contributions. First, we theoretically show that the optimal control policy of heating, ventilation and air conditioning can be well approximated by the threshold policies after natural ventilation is applied; the optimal control policy of natural ventilation can be greatly approximated by the threshold policy, when the indoor air temperature as a function of time has monotonicity after the heating, ventilation and air conditioning policy is applied. Furthermore, we establish a rule-based law framework for the policy approximation of joint control of heating, ventilation, and air conditioning and natural ventilation. Second, we propose the thresholds estimation framework for the policy approximation of fan coil unit, fresh air unit, and natural ventilation respectively, based on the dynamic of the outdoor air temperature, the indoor base air temperature, and the indoor air temperature after the heating, ventilation, and air conditioning policy is applied. Finally, we compare the performance loss, the indoor comfort violation rate, and computational complexity under the approximated policy with those under the dynamic programming, the optimized artificial neural network method, [24–26], and the small sampling and machine learning method [27]. Numerical testing results show that our method saves the computing time dramatically with no effect on the comfort of occupants and little performance degradation.