考虑驾驶风格的心理场效应分析及跟驰行为建模

Psychological field effect analysis and car-following behavior modeling based on driving style

为了更准确地解析车辆间跟驰驾驶行为,充分考虑到驾驶人心理因素的影响,引入心理场理论对车辆的跟驰行为进行建模。研究利用车辆历史轨迹数据,提取车辆动力学指标并使用k-means聚类算法进行驾驶风格分类与判别,得到表征驾驶人驾驶风格的感知系数、反应系数和驾驶风格修正系数并将其引入到心理场中,对心理场效应进行分析,基于全速度差(FVD)模型,建立起考虑驾驶风格的心理场跟驰模型。利用MATLAB软件对不同驾驶风格类型的驾驶人分别进行参数标定和仿真试验,结果表明心理场跟驰模型相比原FVD模型、相互作用势模型能更准确模拟车辆的跟驰行为。

In order to analyze more accurately and model car-following behavior,this study introduces the psychological field theory,taking into consideration the influence of the driver’s psychological factors.The study extracts the vehicle’s dynamics index from its historical trajectory data and then classifies and distinguishes the driving style by using the k-means clustering algorithm.After the perception coefficient,reaction coefficient and driving style correction coefficient that characterize the driving style of the driver are obtained,they are introduced into the psychological field for effect analysis.Finally,based on the Full Speed Difference(FVD)model,a psychological field car-following model is established which takes into account the driver’s driving style.The parameters calibration and simulation experiments of different driving style types are carried out by MATLAB software,and the results show that the psychological field car-following model can more accurately simulate the carfollowing behavior,compared with the FVD model and the interaction potential model.