Research on the prediction of dangerous goods accidents during highway transportation based on the ARMA model

The COVID-19 epidemic has caused a lack of data on highway transportation accidents involving dangerous goods in China in the first quarter of 2020, and this lack of data has seriously affected research on highway transportation accidents involving dangerous goods. This study strives to compensate for this lack to a certain extent and reduce the impact of missing data on research of dangerous goods transportation accidents. Data pertaining to 2340 dangerous goods accidents in the process of highway transportation in China from 2013 to 2019 are obtained with webpage crawling software. In this paper, the number of monthly highway transportation accidents involving dangerous goods from 2013 to 2019 is determined, and the time series of transportation accidents and an autoregressive moving average (ARMA) prediction model are established. The prediction accuracy of the model is evaluated based on the actual number of dangerous goods highway transportation accidents in China from 2017 to 2019. The results indicate that the mean absolute percentage error (MAPE) between the actual and predicted values of dangerous goods highway transportation accidents from 2017 to 2019 is 0.147, 0.315 and 0.29. Therefore, the model meets the prediction accuracy requirements. Then, the prediction model is applied to predict the number of dangerous goods transportation accidents in the first quarter of 2020 in China. Twenty-two accidents are predicted in January, 23 accidents in February and 27 accidents in March. The results provide a reference for the study of dangerous goods transportation accidents and the formulation of accident prevention and emergency measures.