英文摘要 |
This study employs time-series analysis method to set up the ARIMA models to predict accident number, rate, severity for the Sun Yat-Sen freeway. Both monthly and quarterly intervals are used to calibrate the models. For location analysis, the freeway is divided by its travel direction and maintenance districts. 1t is found that the change of accident number and accident severity do follow a monthly and quarterly pattern. Both reach a highest level in July. As for the distribution of accidents, the central district has highest accident number, sum of accident severity, and average severity compared to northern district and southern district. |