Application of artificial neural networks for greenhouse climate modelling

This paper presents the development of nonlinear black-box climate models of typical greenhouses in the Mediterranean area. Using data obtained from actuators and climate sensors in real greenhouses, the problem of neural identification is tackled using a static (non-recurrent) neural network in an autoregressive configuration (NARX). The selection of a set of input variables, a set of input/output vectors for training and a neural structure is included. The relevance of the obtained models is discussed in terms of their potential use for model validation purposes and control.