A study of model predictive control on variable-air-volume air conditioning systems.

摘要

This thesis proposes to apply a model predictive control strategy (MPC) to multiple-zone temperature and ventilation control in view that MPC can deal with MIMO systems and constraints on inputs and outputs systematically. The MPC-based strategy consists of two MPC controllers to handle the fast-response part of the system-AHU, and the relatively slow part-zones respectively. The control objectives are maintaining multiple zone temperatures within a range that the occupants feel thermally comfortable with, and at the same time delivering ventilation air to each zone with a flow rate no lower than the demanded level. The constraint settings are taken advantage of to manipulate the behavior of the MPC controllers in response to different working conditions. The criteria for ventilation control of a multiple-zone recirculating system according to the ventilation rate procedure of ASHRAE Standard 62 are set up and followed by the MPC strategy. A computer-based building simulator is built based on the actual parameters of a real building, and real weather data is used in simulations. A number of simulation-based experiments demonstrate that the MPC strategy achieves the design objectives and is capable of dealing with multiple zone temperature and ventilation control.