Optimized fuzzy control of a greenhouse

Computer systems can be used to control the greenhouse climate in order to improve the culture development and to minimize the production costs. We have a system which allows to acquire the measurements of internal and external temperature and hygrometry, global radiation, and wind velocity. It consists of a heating system, moistening ducts and a static ventilation to control the internal climate. Since 1991, a classical controller on–o1 has been implemented in our experimental greenhouse, which enabled us to have a great number of data 2les. Knowing that the conventional techniques of regulation are di4cult to implement in this type of system (multivariable, nonlinear, nonstationary) where the interdependence of temperature and hygrometry with the meteorological disturbances are strong, we were brought to study the fuzzy controllers. This paper shows that it is possible to successfully control a greenhouse by using these techniques. During the probation period, we compare the various results obtained with these controllers. c © 2002 Elsevier Science B.V. All rights reserved.