The Study of Greenhouse Temperature Modeling Based on Fuzzy Neural Network

The greenhouse micro-climate has many characteristics such as strongly nonlinear, large time delay, strong coupling, strong interference, time-varying. Studying on the modeling and simulation of greenhouse micro-climate is the foundation to establish the ability to automatically control the greenhouse micro-climate and characteristics for our climate of controlled environment agriculture. In this paper, the greenhouse temperature model is establish based on subtractive clustering adaptive fuzzy neural network, using the experimental data to adjust the parameters of the model and parameters can be determined ultimately converge to some value. According to the established model and from the different seasons simulation results in the greenhouse internal temperature, it is proved the temperature model has high accuracy. Using the model to predict the internal temperature of the greenhouse, predicted and measured values also have a good fitting relationship.