NEURAL NETWORK PREDICTION MODELS AS A TOOL FOR AIR QUALITY MANAGEMENT IN CITIES

作者：Skrzypski Jerzy^*; Jach Szakiel Emilia

来源：Environment Protection Engineering, 2008, 34(4): 129-137.

The aim of the study was to examine the possibilities of developing a prognostic methods for the air quality management in cities. The study was focused on the development of the neural network models for predicting the classes of air quality in terms of the daily dust PM(10) concentration. The air quality class was predicted for the following day based on average and maximal daily concentrations. The MLP and RBF models were tested and the results obtained proved to be satisfactory. In the optimal models, false prognoses (in testing series) constituted only 1.9% in the case of predicting average daily concentration and 7.4% in the case of predicting maximum daily concentration. A small prediction error confirmed that neural network models can be an effective tool for the air quality management in cities.

• 出版日期2008