Abstract
In order to realize an optimal control of greenhouse crop irrigation, it is necessary to determine the transpiration. The micro-lysimeter was a good system to sample the transpiration values online but it is an expensive and a complex option, due to maintenance requirements. As an alternative, the use of transpiration models based on climate variables would be a better choice. In this paper, two transpiration models are shown. First, a grey-box static model based on the Penman-Monteith equation was developed, obtaining good results but also some problems were found: different parameter values for different seasons, bad fixing with the lower and higher leaf area index in the estimated transpiration values, and anticipation with respect to the real transpiration in determinate conditions. On the other hand, black-box dynamic models, obtained with system identification techniques, were developed, obtaining better results and solving the problems found in the grey-box model.