LAUSR

As the atmospheric environment pollution has been becoming more and more serious in China, it is highly desirable to develop a scientific and effective early warning system that plays a great significant role in analyzing and monitoring air quality. However, establishing a robust early warning system for warning the public in advance and ameliorating air quality is not only an extremely challenging task but also a public concerned problem for human health. Most previous studies are focused on improving the prediction accuracy, which usually ignore the significance of uncertainty information and comprehensive evaluation concerning air pollutants. Therefore, in this paper a novel robust early warning system was successfully developed, which consists of three modules: evaluation module, forecasting module and characteristics estimating module. In this system, a new dynamic fuzzy synthetic evaluation is proposed and applied to determine air quality levels and primary pollutants, which can be regarded as the research objectives; Moreover, to further mine and analyze the characteristics of air pollutants, four different distribution functions and interval forecasting method are also employed that can not only provide predictive range, confidence level and the other uncertain information of the pollutants future values, but also assist decision-makers in reducing and controlling the emissions of atmospheric pollutants. Case studies utilizing hourly PM2.5, PM10 and SO2 data collected from Tianjin and Shanghai in China are applied as illustrative examples to estimate the effectiveness and efficiency of the proposed system. Experimental results obviously indicated that the developed novel early warning system is much suitable for analyzing and monitoring air pollution, which can also add a novel viable option for decision-makers.