Multi-Model Predictive Control on HVAC Chilled Water Pump for Temperature Control and Energy Consumption Reduction for Auditorium

Most energy consumption comes from the high demand from buildings. A large portion from buildings comes from the HVAC systems. A proper optimal control strategy is needed for energy savings. This paper proposes a multi-model MPC strategy for controlling the chilled water pump HVAC to reduce energy consumption for an auditorium in MAC UI building. The building model is modeled in EnergyPlus software, while the control strategy is modeled using MATLAB where both software communicates using BCVTB. The developed control performance has been tested under different operating conditions and different set of temperature changes. The result shows that the proposed controller can reduce the total energy consumption of chilled water pump by 13.1% compared to the existing On/Off controller.