A modeling approach for suitability evaluation of traffic noise prediction under mixed traffic situation in mid-sized Indian cities

An efficient traffic noise prediction model is a decision-making tool for creating a peaceful environment. Despite the escalating concerns regarding traffic noise in India, limited research focuses on traffic noise prediction to address the unique road traffic scenarios prevalent in the country. Therefore, in the present study, suitability evaluation of the most extensively used prediction models, Federal Highway Administration (FHWA) and Calculation of Road Traffic Noise (CRTN), is carried out. Subsequently, a robust regression model tailored for accurate traffic noise prediction is formulated. However, to modify and enhance the applicability of the model, a novel reference energy mean emission level equation is developed for each vehicle category predominantly plying on Indian roadways. The FHWA approach yielded mean absolute error (MAE) and root mean square error (RMSE) values of 2.5 and 2.6, respectively. In contrast, the CRTN approach exhibited higher errors with MAE and RMSE values of 3.4 and 3.6, respectively. In addition, assessing all other performance metrics demonstrates the better applicability and efficiency of the modified FHWA model with optimum accuracy. The number of heavy vehicles, average traffic speed, and traffic flow were observed to be the vital factors influencing traffic noise generation. Furthermore, using the proposed model results, a regression model for different time durations that can be extensively used in mid-sized Indian cities is established. This research contributes significantly to the field, providing valuable insights for town planners, highway engineers, and authorities to formulate effective noise mitigation strategies in urban planning.