基于VDR数据因子分析的驾驶行为聚类分析

Clustering Analysis of Driving Behavior Based on Factor Analysis of VDR Data

为了解公交车驾驶员的驾驶行为模式,笔者对汽车行驶记录仪(Video Drive Recorder,VDR)保存的数据进行因子分析,得到影响驾驶员行为的速度影响因子、驾车平稳影响因子、起步影响因子与速度控制影响因子4个复合因子,进而对4个复合因子进行聚类,并给出针对不同驾驶员行为的改进方法。分析结果表明,A类、B类驾驶员的驾驶行为低于速度影响因子的平均值,A类驾驶员在驾车平稳影响因子和起步影响因子平均值中高于B类,且C类是大多数驾驶员的驾驶模式。

In order to understand the driving behavior mode of bus drivers, the author makes factor analysis on the data saved by VDR,and obtains four composite factors affecting driver behavior: speed influence factor, driving stability influence factor, starting influence factor and speed control influence factor. Then, the four composite factors are clustered, and the improvement methods for different driver behaviors are given. The analysis results show that class A The behavior of type B drivers is lower than the average value of speed influence factor,while the average value of smooth driving influence factor and start influence factor of type A drivers is higher than that of Type B, and the driving behavior of type C is the driving mode of most drivers.