Soft sensing of plant shoot-tip temperature

A neural network model which predicted plant shoot-tip temperature was constructed on six crops. The model had three layers, as input, middle and output layers, and the inputs of the model were drybulb temperature, wetbulb temperature, glazing temperature and solar radiation in greenhouse, then the output was shoot-tip temperature. The data for training and verification were collected in greenhouse over several months. The simulation study was performed with the constructed model, and it was found that the predicted shoot-tip temperature closely agreed with the measured data and the model was applicable as a tool to monitor plant shoot-tip temperature in commercial greenhouses.