MPC for environmental control of a vertical farm

Abstract

Horticulture environmental control has a high impact on growers profits in enclosed growth environments. The following thesis explores two well known control algorithms, the PID and MPC to control air temperature, carbon dioxide and humidity within a vertical farm. Both controllers were tuned using a developed nonlinear model and are compared in performance with respect to economics. Controllers were simulated for 1 day and once tuned, deployed in an experimental system for 3 days. The MPC had better performance in air temperature and humidity compared to the PID, although in the case of air temperature it used more energy. Both PID and MPC struggled with carbon dioxide control, and at the end of both experiments, the biomass was similar. Both controllers had the capability of following the reference trajectories and being applicable in enclosed growth environments.