Search

Air pollution in metropolises is one of the serious problems of human life. Tehran is one of the cities facing air pollution problem. Urban managers concern about choosing different management methods to control air pollution. In this study, a combination of fuzzy systems and neural networks has been used to select the most suitable scenario for controlling SO 2 pollution. According to the method presented in this paper, 8 input data categories such as wind speed, precipitation, temperature, pressure, humidity, gas oil consumption, gasoline consumption and urban green space levels have been used as independent parameters and SO 2 pollutant concentration has been considered as the dependent parameter. The contribution of each meteorological station to the meteorological data was determined by Thiessen Polygon Method. Then, using adaptive neural fuzzy inference systems, modeling was done in Sugeno Method and the least root mean square error (3.19) was determined for the model. Then, by changing each of the independent parameters, the effect of each of these independent parameters on SO 2 pollutant was measured. The results showed that the parameters of pressure, urban green space, gasoline consumption, gas oil consumption, temperature, wind speed and humidity, respectively, had the greatest effect on reducing the SO 2 concentration. Since the parameters of gasoline and gas oil consumption as well as the area of green space are changeable by different policies and by human decisions, the concentration of SO 2 pollutant can be controlled by reducing the consumption of gasoline and gas oil and increasing the green space in Tehran. • Finding the lowest error rate in any of the models predicting sulfur dioxide concentration. • Determine the effect of each of the independent parameters on the concentration of SO 2. • Determination of the best predictive model for sulfur dioxide emissions based on ANFIS model. • Providing solutions to improve air pollution. [ABSTRACT FROM AUTHOR]