ABSTRACT: Unpleasant sound is known to be a solemn menace to life and social existence, the relevance of this research work is in the application of neural networks to real-time traffic data from road traffic sound measurement. Traffic data were collected with the aid of a sound meter of a high accuracy grade. The peak traffic sound level that causes disturbance to human beings known as L10 was measured in (dB(A)) which was sampled in Twelve (12) locations selected for the study in Lokoja, Kogi State Nigeria, and the associated noise levels were predicted by the application of a neural network. The output obtained shows the highest value of L10 to be 86 dB(A) while the lowest value is 73 dB(A) in the sites considered. The results of the Neural Network output show a strong connection between the results obtained when the calculation of the road traffic noise model (CRTN)
a methodology was deployed.
Keywords: Neural Network; Noise pollution; Multi-layer perceptron; L10.
JoST. 2021. 11(1): 26-31
Accepted for Publication, April 25, 2021
