PID-MPC implementation on a chiller-fan coil unit

Abstract

A proportional, integral, and derivative-model predictive controller (PID-MPC) is implemented to enhance the performance of a chiller-fan coil unit with regards to temperature, humidity, and CO2 level control. The sublayer PID is implemented for outdoor air processing to level out fresh and room air temperatures. Precooled outdoor air prevents from imbalanced loads due to fresh air intake that could delay set point tracking and increase the compressor and supply fan speeds. The upper layer MPC is implemented to control temperature, humidity and CO2 level simultaneously. The coupling effect between temperature and humidity is considered, whilst an optimal control action is determined. The MPC is also implemented to withstand disturbances. The PID performance was satisfactory in terms of settling time and maximum overshoot. The MPC performance was satisfactory in terms of set point tracking and disturbance rejection.