PREDICTION OF GREENHOUSE MICRO-CLIMATE USING ARTIFICIAL NEURAL NETWORK

The aim of this study is to develop an Artificial Neural Network (ANN) model for prediction of one day ahead mean air temperature and relative humidity of greenhouse located in the sub-humid sub-tropical regions of India. The adequacy of back propagation neural network to model the inside temperature and humidity of a production greenhouse as a function of micro-climatic parameters including temperature, relative humidity, wind speed, and solar radiation was addressed. Micro-climatic data of greenhouse and outside were collected on daily basis and used for analysis of best fit ANN model. After the network structure and parameters were determined reasonably, the network was trained. The activation functions were respectively the hyperbolic tangent in the hidden layer and the linear function in the output layer. The Root Mean Square Error (RMSE), Mean Absolute Error (MAE) and Correlation Coefficient were chosen as the statistical criteria for measuring of the network performance. A comparison was made between measured and predicted values of temperature and relative humidity, and the results showed that the BP neural network (for network (6-4-2) model given a best prediction for inside temperature and relative humidity. Statistical analysis of output shows that, the RMSE and Mean Absolute Error (MAE) between the measured and predicted temperature was 0.711 degrees C and 0.558 degrees C, and the relative humidity RMSE and MAE was 2.514% and 1.976% which can satisfy with the demand of greenhouse climate control.