Automatic Synthesis of Greenhouse Supervisory Climate Control with Integrated Crop Growth Models

In recent years, automated climate control has been applied in greenhouses to achieve higher yields, better quality, and more efficient resource use. However, this control methodology simply adapted traditional building climatic control to the greenhouse context, without accounting for the effects of the plants within. In this paper, we transfer advanced modeling and control methodologies from systems engineering to maintain optimal temperature and CO2 levels for plant growth. In particular, we introduce hybrid automata and supervisory control theory to automatically synthesize controllers that not only account for the usual greenhouse thermal effect, but also for the presence of plants, whose CO2 absorption rate changes over time due to daylight-dependent photosynthesis and canopy growth. Simulation results confirm that the synthesized controllers maintain temperature and carbon fertilization level within the desired bounds.