Research On Air Quality Index Of Jinan Based On ARIMA Composite Forecasting Model

In 2016,a lot of cities in our country are shrouded in smog,and the haze is no longer a beauty,and the smog is a huge obstacle to normal working life.In recent years,the smog has been prolonged and the concentration of pollutants continues to rise,and the problem of air quality has become a problem.According to the change rule of the air quality index,forecasting the future development trend is the key to the problem of air pollution control and governance.Initially,commonly used time series analysis is a single model,but a single model has its own limitations.Combination forecast model to forecast that use a variety of sample information larger limit,than a single prediction model more systematic,more comprehensive.Then,many scholars began to study combination forecast,and they applies a composite forecasting model to the analysis and forecast of air quality data.They shows that the method is used to study the feasibility of air quality issues.Under the guidance of this thought,I studied the optimal combination forecast model and variable weight combination forecast model.This paper attempted to establish the variable weight combination forecasting model based on cumulative residuals.Residual is the objective reflection of prediction accuracy.This article uses the cumulative residual to structure weight coefficient Weight shows that the importance of all kinds of forecasting method to the model.This makes the model more adapt to the change trend of the data,so as to achieve more ideal,more robust fitting prediction effect.Monitoring data of the jinan’s air quality index is applied to the model,from theoretical and practical application of two direction confirms the feasibility of this method,the validity and reliability,and the analysis for the forecast of air quality index has certain reference significance.