In the issue of expanding noise levels the worldover, road traffic noise is main contributor. The investiga-tion of street traffic noise in urban communities is a sig-nificant issue. Ample opportunity has already passed tounderstand the significance of noise appraisal through pre-diction models with the goal that assurance against streettraffic noise can be actualized. Noise predictions models areutilized in an increasing range of decision-making applica-tions. This study’s main objective is to assess ambient noiselevels at major arterial roads of Surat city, compare thesewith prescribed standards, and develop a noise predictionmodel for arterial roads using an Artificial Neural Network.The feed-forward back propagation method has been usedto train the model. Models have been developed using thedata of three roads separately, and one final model has alsobeen developed using the data of all three roads. Amongthe prediction in three arterial roads, the predicted outputresult from the model of Adajan-Rander showed a bettercorrelation with a mean squared error (MSE) of 0.789 andR2value of 0.707. But with the combined model, there isa slight deterioration in mean squared value (MSE) 1.550,with R2not getting changed much significantly,i.e., 0.755.However, the combined model’s prediction can be adopteddue to the variety of data used in its training.
Development of traffic noise prediction model for major arterial roads of tier-II city of India (Surat)using artificial neural network
