Nonlinear Set Membership forecast of urban ozone peaks

In the paper a Nonlinear Set Membership (NSM) prediction method isapplied to forecast tropospheric pollution. The NSM method doesnot require the choice of the functional form of the model usedfor prediction but only assumes a regularity condition on theregression function defining the model. In this way, thecomplexity/accuracy problems deriving from the proper choice of asuitable parametrization are circumvented. Here the NSM method isused to forecast the tropospheric ozone concentration in Brescia,a highly populated and industrialized area in Northern Italy.Purpose of this application is to aid local Authorities indecision-making policies for secondary pollution control andprevention. The NSM method is compared to other modellingapproaches such as Neural Network, Neuro-Fuzzy, ciclostationaryautoregressive and k-nearest neighbor classification. Modelperformances are assessed through comparison of indices andstatistical indicators suggested by the European EnvironmentAgency.