System identification: theory for the user
System identification: theory for the user
Optimized fuzzy control of a greenhouse
Fuzzy Sets and Systems - Featured Issue: Selected papers from ACIDCA 2000
Greenhouse air temperature predictive control using the particle swarm optimisation algorithm
Computers and Electronics in Agriculture
| Hi-index | 0.01 |
This paper deals with the problem of modelling and control of greenhouses inside climate defined by two variables: the temperature and hygrometry. The control objective aims to ensure a favourable inside microclimate for the culture development and to minimize the production cost. Achieving this objective is difficult, due to the complexity of the phenomena involved in the plant growth process: the two variables are correlated and very sensitive to the outside weather and also to many other practical constraints (actuators, moistening cycle ...). We propose highly performing regulation for the greenhouse internal state based on H"2 robust control design. It involves a linear control model of the process, obtained by an off-line parametric identification technique. Evaluation of control performance is achieved through a benchmark physical model derived from energy balance for the temperature and water mass balance for the hygrometry. The main steps in deriving this nonlinear model are also outlined. A successful feasibility study of the proposed controller is presented for an experimental greenhouse located at the University of South Toulon-Var (France). Simulation results show promising performances despite the high interaction between the process internal variables and the high impact on these variables of the external meteorological conditions.