R Research On Traffic Accident Problem Based On ARIMA-EGARCH Model

This paper analysis traffic accident incidents to reduce the direct losses and casualties.Xie[1]and Li[2]studied the traffic accidents by establishing ARM A and ARIMA models,analyzed data on deaths and mortality rates,respectively,and then proceeded to the next step.Road traffic system is a dynamic time-varying system.The factors affecting traffic safety are complex and changeable.Many fac-tors are grey.Xie[3]predicted the number of accidents,deaths,injuries and direct economic losses based on Grey System Theory.However,when the range of data change is too large,the grey model will have the problems of poor coincidence and low accuracy.In this paper,time series method is considered.Taking the national traffic accident data from 1990 to 2017 as the research object,ARIMA(0,1,4)model is established through data preprocessing and model identification,verification and optimization.Xie[1]and Li[2]did not consider the heteroscedasticity,and the test found that the residual sequence had heteroscedasticity.In this paper,the EGARCH(1,1)model was established to eliminate the heteroscedasticity,so that the prediction results were more accurate.At the same time,several models such as exponential smoothing,moving average and multiple linear regression are estab-lished in this paper.Finally,the data from 2013 to 2017 are used as assessment samples to test the accuracy of the model.The results show that the prediction error of ARIMA-EG ARCH model is smaller,so this model can be used to predict the data from 2018 to 2019.