Prediction and estimation of atmospheric pollutant levels by soft computing approach

The goal of this paper is to design an environmental monitoring system which is able to estimate and predict the pollutant values of an important city area on the Strait of Messina (Italy). In order to solve the problem, we propose the use of neuro-fuzzy inference techniques. This approach utilizes the concepts of fuzzy inference systems (FIS) to estimate and predict the pollution level in the air. By using a specific MatLab^R Toolbox, we developed a sophisticated FIS. Each rule is of the IF...THEN structure in terms of linguistic framework in which the easy understanding due to the open box structure can help the politicians to take decisions about the urban traffic