Fuzzy subspace-based constrained predictive control design for a greenhouse micro-climate

The efficiency of greenhouse production relies mainly on the quality of its interior micro-climate. However, abrupt changes in the indoor climate conditions can cause damage to the plants. The climate control of greenhouses is becoming a viable solution for creating a favorable environment for plants growth. In the present work, a data-driven identification-based model predictive control scheme is developed to track the desired environmental conditions inside the greenhouse while satisfying the physiological constraints of the crops. Where a fuzzy subspace state-space system identification technique is performed to identify the greenhouse system. A linear model predictive controller subject to both input-output magnitude and rate of change constraints is designed to enable adequate temperature and humidity tracking with required characteristics. A comparative study of the proposed approach with nonlinear model predictive control is presented. The obtained results have shown its feasibility with respect to tight climate quality constraints and advantageous performances in terms of computational cost.