基于多维数据的公交驾驶员安全风险评价及诊断

Risk Assessment and Diagnosis of Bus Drivers Based on Multi-Dimensional Data

以多维数据为驱动,实现个体驾驶员完整数据链条,并建立多维度公交驾驶员风险评价指标体系。在此基础上,构建基于熵权-TOPSIS法和LightGBM法的风险评价模型,通过随机森林分维度提取并解析个体驾驶员风险致因。研究成果为全面、高效、客观地评价驾驶风险提供了新思路,同时也为推动公交事业数字化进程提供了更高效、更精准的安全管理手段。

Driven by multi-dimensional data,this paper realizes the complete data chain of individual drivers,and establishes a multi-dimensional risk evaluation index system for bus drivers.On this basis,a risk assessment model based on entropy weight-TOPSIS method and LightGBM method was constructed,and the risk causes of individual drivers were extracted and analyzed through random forest dimensions.The research results provide a new idea for comprehensive,efficient and objective evaluation of driving risks,and also provide a more efficient and accurate safety management means for promoting the digitalization of public transportation.