In the present work, vehicular traffic noise prediction models have been developed for Patiala city (Punjab) using GA and regression approach. The various terminologies related to GA and acoustics analysis are discussed. The models predict equivalent continuous sound level (Leq) as the function of vehicle volume (Log Q) and percentage of heavy vehicles (P%). A large number of data have recorded at different dates/timings to account variability. Three commonly used GA selection operators (uniform, roulette wheel, and tournament) are used to analyse the accuracy of GA models. The GA model performs better as compared to regression model. The average mean square error (MSE) using GA model is 0.59 as compared to 0.76 for regression model. Among all GA selection operators, tournament selection shows better result.
Regression, Genetic Algorithm, Modelling, traffic noise, vehicle volume, percentage of heavy vehicles.
Paras Kumar, Traffic Noise Estimation using Genetic Algorithm (GA) Approach, HCTL Open International Journal of Technology Innovations and Research (IJTIR), Volume 16, July 2015, e-ISSN: 2321-1814, ISBN: 978-1-943730-43-8.