Vehicular traffic noise prediction using soft computing approach.
J Environ Manage
; 183: 59-66, 2016 Dec 01.
Article
em En
| MEDLINE
| ID: mdl-27576153
ABSTRACT
A new approach for the development of vehicular traffic noise prediction models is presented. Four different soft computing methods, namely, Generalized Linear Model, Decision Trees, Random Forests and Neural Networks, have been used to develop models to predict the hourly equivalent continuous sound pressure level, Leq, at different locations in the Patiala city in India. The input variables include the traffic volume per hour, percentage of heavy vehicles and average speed of vehicles. The performance of the four models is compared on the basis of performance criteria of coefficient of determination, mean square error and accuracy. 10-fold cross validation is done to check the stability of the Random Forest model, which gave the best results. A t-test is performed to check the fit of the model with the field data.
Palavras-chave
Texto completo:
1
Coleções:
01-internacional
Base de dados:
MEDLINE
Assunto principal:
Modelos Teóricos
/
Ruído dos Transportes
Tipo de estudo:
Prognostic_studies
/
Risk_factors_studies
País/Região como assunto:
Asia
Idioma:
En
Ano de publicação:
2016
Tipo de documento:
Article