卫星定位数据驱动的营运车辆驾驶人驾驶风险评估模型

为了提高营运车辆驾驶人安全管理的精细化水平，合理地评估驾驶人驾驶风险程度，有的放矢地降低高风险驾驶人的事故率，基于卫星定位数据特点及驾驶行为与驾驶风险的相关关系设计26个驾驶行为特征参数。考虑到高速和非高速行驶时相同驾驶行为对驾驶风险的影响区别较大，根据23名营运车辆驾驶人的实测数据有针对性地筛选高速和非高速路段驾驶人风险评估指标，构建营运车辆驾驶人驾驶风险评估指标体系。然后，基于熵权法、独立性权系数法和Spearman相关系数法建立集成赋权法，确定各评估指标的权重。最后，雇佣40名营运车辆驾驶人进行实车试验以验证模型的合理性。结果表明：车辆速度和加速度方面的驾驶行为特征可以用于评估驾驶人的驾驶风险且评估效果较好，驾驶风险评估得分与实际交通冲突次数呈正相关关系，所建立模型可以较为准确地评估营运车辆驾驶人驾驶风险的高低，准确率达到77.50%，该模型在不同地区使用时，准确率存在一定的差异，但在容许范围之内，方法具有较好的鲁棒性。

Driving Risk Assessment Model of Commercial Drivers Based on Satellite-positioning Data

To improve the safety of commercial drivers, assess the driving risk level of drivers, and reduce the accident rate of high-risk drivers. Subsequently, 26 driving behavior characteristic parameters were designed based on the satellite-positioning data and correlation between driving behavior and driving risk. Considering that the influence of the driving behavior on driving risk is different between high-speed and low-speed driving, we used real-time data of 23 commericial drivers on an on-road driving test to screen driving risk assessment indicators and developed the driving risk assessment indicator system of commercial drivers. Then, the integrated weighting method was established based on the entropy weight, independent weight coefficient, and Spearman correlation coefficient methods to determine the weight of each indicator. Finally, real vehicle verification tests were performed with 40 commercial drivers to verify the rationality of the proposed model. The results show that the driving behavior in terms of speed and acceleration can assess the accident risk, and the effect is good. The risk assessment score of the assessment model is positively correlated with the actual number of traffic conflicts. The established model can assess the driving risk level of commercial drivers with an accuracy rate of 77.5%. The accuracy of the model varies based on locations, but the model exhibits better robustness within a specified range.