Application of Artificial Intelligence for Fuel Consumption Prediction of a Tractor in Different Operating Conditions

Artificial neural networks (ANNs) have been included in the field of agriculture time after time because of its benefits over conventional frameworks. Prediction and forecasting based on parallel reasoning is the primary advantage of neural networks. The present study was conducted to develop a fuel consumption prediction model for N 4710 tractor in any operating conditions. Different types of neural network models were prepared. Among them, the radial basis function neural network was selected as the best neural network based on the coefficient of determination (R
²
) value and mean squared error. Coefficient of determination (R
²
) of 0.99 for training and 0.95 for testing datasets was observed with a mean squared error of 0.01 for both training and testing datasets. It was found from the results that; the fuel consumption is proportional to the engine power at varying load and throttle settings. The developed neural network can predict the fuel consumption of a tractor in the given power range under various load and throttle settings with significant accuracy (MAE = 0.05). The model was validated by comparing the predicted and actual fuel consumption data of a selected tractor model N 4710 tested by the FMTTI India. The coefficient of determination (R
²
) was 0.90 and MSE was 0.004 for the predicted fuel consumption by the developed ANN at varying load and throttle settings (100% and 50%).