Mohamad Reza Moghimi, Mohsen Pooya, Behzad Mohammadi Alasti and Mehdi Abasgholipor Ghadim
The aim of this study was to examine energy use pattern and predict the output energy for dry wheat production in Gorve country, Kordestan province of Iran. The data used in this study were collected from farmers by using a face to face survey. The results revealed that wheat production consumed a total of 42.998 G J ha–1 and output was 97.935 G J ha–1. Electricity has the highest share by 26.135 G J ha–1 followed by total fertilizers and diesel fuel. In this study, several direct and indirect factors have been identified to create an artificial neural networks (ANN) model to predict output energy for wheat production. The final model can predict output energy based on human labor, machinery, diesel fuel, chemical fertilizer, biocides, electricity and seed. The results of ANNs analyze showed that the (7-6-6-1)-MLP, namely, a network having six neurons in the first and second hidden layer was the bestsuited model estimating the output energy. For this topology, MSE and R2 were 0.003 and 94%, respectively. The sensitivity analysis of input parameters on output showed that total chemical fertilizer and seed had the highest and lowest sensitivity on output energy with 22% and 7%, respectively.
