Using ANN and EPR models to predict carbon monoxide concentrations in urban area of Tabriz

Mohammad Shakerkhatibi1*, Nahideh Mohammadi2, Khaled Zoroufchi B, (2015) Using ANN and EPR models to predict carbon monoxide concentrations in urban area of Tabriz. Environmental Health Engineering and Management Journal.

[img]
Preview
Text
117-122.pdf - Published Version

Download (1MB) | Preview

Abstract

Background: Forecasting of air pollutants has become a popular topic of environmental research today. For this purpose, the artificial neural network (AAN) technique is widely used as a reliable method for forecasting air pollutants in urban areas. On the other hand, the evolutionary polynomial regression (EPR) model has recently been used as a forecasting tool in some environmental issues. In this research, we compared the ability of these models to forecast carbon monoxide (CO) concentrations in the urban area of Tabriz city. Methods: The dataset of CO concentrations measured at the fixed stations operated by the East Azerbaijan Environmental Office along with meteorological data obtained from the East Azerbaijan Meteorological Bureau from March 2007 to March 2013, were used as input for the ANN and EPR models. Results: Based on the results, the performance of ANN is more reliable in comparison with EPR. Using the ANN model, the correlation coefficient values at all monitoring stations were calculated above 0.85. Conversely, the R2 values for these stations were obtained <0.41 using the EPR model. Conclusion: The EPR model could not overcome the nonlinearities of input data. However, the ANN model displayed more accurate results compared to the EPR. Hence, the ANN models are robust tools for predicting air pollutant concentrations. Keywords: Forecasting, ANN, EPR, Carbon monoxide, Modeling

Item Type: Article
Subjects: G Geography. Anthropology. Recreation > GE Environmental Sciences
Depositing User: ehemj ehemj ehemj
Date Deposited: 15 Dec 2015 07:06
Last Modified: 15 Dec 2015 07:06
URI: http://eprints.kmu.ac.ir/id/eprint/22194

Actions (login required)

View Item View Item