Prediction of agricultural tractor noise levels using artificial neural networks

Agricultural tractors generate noise pollution in the cabin and in open air. The demands for good sound comfort of the driver inside cabin and assistant driver in outside of these tractors are continuously growing. The main objective of this paper is to predict the noise levels surrounding the tractor operator and in open air by using Artificial Neural Networks (ANNs) and to compare the results against noise levels from collected data. The architecture of the network is used with the backpropagation algorithm - the multilayer feedforward networks. Another objective is to predict the tractor noise levels at various operating speeds and determine their influences on the surrounding noise. The use of a tractor may be avoided in the even its noise level exceeded safe levels. This can be actuated by a specific control system that selects the optimum speed generating safe noise levels.