Midterm Air Pollution Monitoring and Prediction Based on Adaptive Neural Fuzzy Inference System

The prediction air pollutants plays a decisive role in taking preventive measures in the community. In this paper, using the data received from the measurement centers of the air pollutants in different regions of Tehran and with the help of the intelligent approach of the adaptive networks-fuzzy inference systems, a scheme is designed that can automatically Forecast the amount of air pollutants in a few hours in future. For this purpose, two common pollutants of Sulfur dioxide and particulate matters have been selected. Therefore, in short-term and mid-term phases, the proposed algorithm is studied and the simulation results are analyzed in each phase. Simulation results indicate high accuracy in the short term forecast as well as acceptable accuracy in the mid-term forecast.