The Forecast Of AQI Based On Time Series Analysis

The air quality effects people’s physical and mental health and their daily life. The index of air quality is an very important index to evaluate the air quality,and it puts some kinds of main air pollution which impact the health of people into the form of single index value.It can be used to judge the air quality. The higher it is, the more serious the air pollution is. Using AQI to evaluate the air quality can accurately response the actual situation.It is a most commonly used index for measuring air quality at present. The result of it is more close to the real feeling of people.We use curve fitting and parameter estimation to establish a time series analysis fitting model base on the characteristics of time series.Not only can it reveal the inherent law of everything through the digital rule, but also it can predict and control the development of things.It has strong practicability in economic,trade,business and other areas of our life.It is also a kind of important method to predict air quality.This paper makes the daily data of AQI of dalian in 2014 to be a sample,and uses Eviews software and time series analysis to establish fitting models.We forecast that the time series is a stationary series, then use AR(p) model, MA(p)model and ARMA(p,q) model to predict. Through the analysis and comparison,we find the predicted result of ARMA(1,1) model is better than the others.Therefore, we chose it as the fitting model of this series.Finally,we use this model to forecast the air quality index ten days after in dalian. The result shows what level the dalian air quality is, and whether it is good and suitable for people’s daily life.