作者: P. J. García Nieto , F. Sánchez Lasheras , E. García-Gonzalo , F. J. de Cos Juez
DOI: 10.1007/S00477-018-1565-6
关键词: Autoregressive model 、 Mathematical model 、 Air quality index 、 Particulates 、 Pollutant 、 Meteorology 、 Autoregressive integrated moving average 、 Environmental science 、 Support vector machine 、 Metropolitan area
摘要: Atmospheric particulate matter (PM) is one of the pollutants that may have a significant impact on human health. Data collected over 7 years from air quality monitoring station at LD-III steelworks, belonging to Arcelor-Mittal Steel Company, located in metropolitan area Aviles (Principality Asturias, Northern Spain), analyzed using four different mathematical models: vector autoregressive moving-average, integrated moving-average (ARIMA), multilayer perceptron neural networks and support machines with regression. Measured monthly, average concentration (SO2, NO NO2) PM10 (particles diameter less than 10 μm) used as input forecast monthly 7 months ahead. Simulations showed ARIMA model performs better other models when forecasting 1 month ahead, while 9 months ahead best performance given by