An observer based backstepping control for a greenhouse

摘要

In this paper we use the backstepping approach to control climate parameters of greenhouse. To get realistic simulation, we use real weather data collected in the south region of Algeria (Biskra). We control temperature, relative humidity and CO2 concentration by acting on four variables such as heating, ventilation, CO2 injection and water injection. However as the CO2 sensors is not available, we use an nonlinear observer based on temperature and steam density measurements, To get an optimal controller, the multi- objective genetic algorithms based on Non dominated Sorting Genetic Algorithms (NSGA) techniques and Multi-Criteria Decision Analysis (MCDA) approach are introduced.