Fuzzy Model based Prediction of Ground-Level Ozone Concentration

Abstract

Ground-level ozone is a dangerous pollutant for which the prediction of the concentration could be of great importance. In this paper, we present and compare three fuzzy models aiming the forecasting of ground-level ozone concentration. The models apply Takagi-Sugeno, respective LESFRI fuzzy inference techniques and were generated using the ANFIS method of the Matlabs Fuzzy Logic ToolBox, respective the RBE-DSS method of the SFMI toolbox. Although all of the methods proved to be applicable the model using LESFRI ensured the best results with a low number of rules.