Traffic noise modelling using genetic algorithm

Abstract

The major contribution of the traffic noise, towards overall noise pollution scenario is well known established fact. Traffic noise from highways creates problems for surrounding areas, especially when there are high traffic volumes and high speeds. Vehicular traffic noise problem is contributed by various kinds of vehicles like heavy,medium trucks/buses, automobile and two wheelers. In India, the transportation sector is growing rapidly and number of vehicles on Indian roads is increasing at very fast rate. This has lead to overcrowded roads and noise pollution. So, a need is being felt to develop a noise prediction model suitable for Indians roads condition. A mathematical model is developed in Patiala city (Punjab) for a site at Patiala- Chandigarh highway (10 km from Thapar University). A large number of sets of data were recorded for 15 mins duration. The noise measurement parameters to be recorded were Leq (15 min). Sound level meter (CESVA SC 310) was used for these measurements. The Developed mathematical model which can be used for predicting Leq, level included the parameters such as Total vehicle volume/hr, percentage of heavy vehicles, and average speed of vehicles. The Noise levels Leq is used in Genetic algorithm and regression analysis for prediction. It was concluded that value of R2 ranges from 0.88 to 0.99. The paired t-test was also carried out successfully for goodness-of-fitness.