Expert control based on neural networks for controlling greenhouse environment

Abstract

Depending upon the nonlinear feature between neural units in artificial neural networks (ANN), artificial neural network was used to develop a model for greenhouse inside air temperature management. Data was collected and processed for training and simulation of temperature model. The design of the network structure and transfer function was also discussed. Because heuristic logic of greenhouse environment exists, a simple ANN controller does not perfect complex task, the algorithm of ANN with heuristic logic is developed. Experimental result indicated that the performance of the new control scheme meet the requirement of greenhouse environment control. © 2010 IFIP International Federation for Information Processing.