Application of artificial neural network to prediction of traffic noise levels in the city of Niš, Serbia

Abstract

© 2018, Springer International Publishing AG. Road traffic noise is one of the main sources of environmental pollution in urban areas. Since the early 1950s many mathematical models for traffic noise prediction have been developed. Available models are usually based on regression analysis of experimental data. This paper presents the application of multilayer feedforward neural network in prediction of equivalent continuous level of road traffic noise in urban areas of the city of Niš, Serbia. Predictions of developed neural network are compared to data collected by traffic noise monitoring, as well as to predictions of commonly used traffic noise models. Obtained results show that the application of artificial neural network may improve accuracy of traffic noise prediction.