Evaluating the Accident Risk Level of Truck Drivers Based on Aberrant Driving Behavior

抄録

Historical statistics show that the number of accidents involving trucks is not many compared to those involving other vehicle types, yet the crash rate and the casualty rate of truck crashes are much higher than those of sedans. Truck crashes also cause huge financial losses and traffic jams. Therefore, truck companies must manage drivers by implementing cost-effective measures. This study proposed a revised approach based upon previous studies assessing driving risk of commercial vehicles. Combining Jenks natural breaks optimization and fuzzy logic has increased the logicality and flexibility of the evaluation system. Moreover, to illustrate the impact of each driving behavior on the risk level from the perspective of decision-making, this study established decision tree models. These tree models provide a quick and visible tool to grade the driving risk levels of drivers.