Fuzzy Time Series for Forecasting Pollutants Concentration in the Air

Abstract

The main link from Europe to Sicily and vice versa is represented by a ferry boat service from Villa San Giovanni to Messina, two towns along the homonymous strait. Consequently, the crossing and heavy vehicular traffic discharges organic and inorganic compounds in the urban atmosphere that can affect the health of citizens. In this paper, a design of a prediction system for pollutants concentration in the urban atmosphere is proposed. In this study, the exploited tool is a soft computing approach basing on Fuzzy Time Series. In particular, we exploit Two Factors Fuzzy Time Series in which the main factor is the pollutant and the secondary factor is the traffic. For our purposes, two algorithms are exploited. The obtained predictions have given good results especially regarding the prediction of pollutant concentrations one hour later, which represents the time needed to take decisions about the traffic regulation.