Prediction of Power Tiller Noise Levels Using A Back Propagation Algorithm

The use of neural networks methodology is not as common in the investigation and predictionnoise as statistical analysis. the application of artificial neural networks for predictionof power tiller noise is set out in the present paper. the sound pressure signals fornoise analysis were obtained in a field experiment using a 13-hp power tiller. duringmeasurement and recording of the sound pressure signals of the power tiller, the enginespeeds and gear ratios were varied to cover the most normal range of the power tiller operationin transportation conditions for the asphalt, dirt rural roads, and grassland. signalsrecorded in the time domain were converted to the frequency domain with the help ofa specially developed fast fourier transform (fft) program. the narrow band signalswere further processed to obtain overall sound pressure levels in a-weighting. altogether,48 patterns were generated for training and evaluation of artificial neural networks. artificialneural networks were designed based on three neurons in the input layer and oneneuron in the output layer. the results showed that multi layer perceptron networks witha training algorithm of back propagation were best for accurate prediction of power tilleroverall noise. the minimum rmse and r2 for the four-layer perceptron network with asigmoid activation function, extended delta-bar-delta (ext. dbd) learning rule withthree neurons in the first hidden layer and two neurons in the second hidden layer, were0.0198 and 0.992, respectively.