2006
Feng, T.; Yang, Z.; Zhang, J.; Fujiwara, A.
Modeling pollutant concentration based on artificial neural network Proceedings Article
In: 2006, ISBN: 7030174445.
Abstract | BibTeX | Tags: Artificial neural network, Meteorological conditions, Pollutant concentration, Traffic flow
@inproceedings{Feng2006,
title = {Modeling pollutant concentration based on artificial neural network},
author = {T. Feng and Z. Yang and J. Zhang and A. Fujiwara},
isbn = {7030174445},
year = {2006},
date = {2006-01-01},
journal = {Proceedings of the Conference on Traffic and Transportation Studies, ICTTS},
abstract = {The traditional two-stage (i.e., emission and dispersion) model, which evaluates air pollutant concentration on roadsides, calculates first the emission from cars, and then estimates the concentration levels at surrounding areas. Due to the complexity of cause-effect relationship between traffic flows and pollutant concentration, it is quite difficult to completely obtain data required for the model. This study develops a new prediction model of vehicular pollutant concentration on roadsides using artificial neural network (ANN) approach. The model not only incorporates the mechanism of pollutant emission and dispersion, but also represents the effects of multiple factors which are expected to influence vehicular pollutant emission and concentration, such as meteorological parameters, traffic flow parameters and road space characteristics. The proposed model can predict the pollutant concentration at any points on roadside with a one-step calculation method by considering the above complicated relation. A large set of survey data is used to train and simulate the ANN-based model. The effectiveness of the proposed model is empirically confirmed.},
keywords = {Artificial neural network, Meteorological conditions, Pollutant concentration, Traffic flow},
pubstate = {published},
tppubtype = {inproceedings}
}
The traditional two-stage (i.e., emission and dispersion) model, which evaluates air pollutant concentration on roadsides, calculates first the emission from cars, and then estimates the concentration levels at surrounding areas. Due to the complexity of cause-effect relationship between traffic flows and pollutant concentration, it is quite difficult to completely obtain data required for the model. This study develops a new prediction model of vehicular pollutant concentration on roadsides using artificial neural network (ANN) approach. The model not only incorporates the mechanism of pollutant emission and dispersion, but also represents the effects of multiple factors which are expected to influence vehicular pollutant emission and concentration, such as meteorological parameters, traffic flow parameters and road space characteristics. The proposed model can predict the pollutant concentration at any points on roadside with a one-step calculation method by considering the above complicated relation. A large set of survey data is used to train and simulate the ANN-based model. The effectiveness of the proposed model is empirically confirmed.