基于实测数据的交叉路口广义力模型改进  

Improvement of GF model in intersections based on actual data

作　　者：张蓉蓉[1] 郑永安[1] 史忠科[1] 

机构地区：[1]西北工业大学自动化学院,陕西西安710129

出　　处：《系统工程学报》2016年第1期111-116,共6页Journal of Systems Engineering

基　　金：国家自然科学基金重点资助项目(61134004)

摘　　要：为了更准确的描述道路交叉路口车辆的跟驰行为,在广义力模型的基础上,提出了改进模型.改进模型去掉了Heaviside函数的影响,添加了两车速度差和车间距的高次项对后车加速度影响的补偿项.实验所用的数据是通过Matlab软件处理在西安市边家村实际道路交叉路口所拍摄的车辆跟驰视频中获得的.用实测数据分别对广义力模型和改进模型进行仿真验证实验并对实验结果进行了对比分析,结果表明改进模型能更好地描述实际道路交叉路口的车辆跟驰行为.In order to describe the car-following behavior of the intersections more accurately, on the basis of generalized force（GF） model, an improved model was proposed. Compared with the GF model, the improved model removed the influence of Heaviside function and added the influence of the higher order terms of speed difference and vehicle distance on the following car＇s acceleration as the compensation. The real-measured data used in the experiment were obtained by Matlab software from the video taken at the intersection near Bian Jia village in Xi＇an City. The real-measured data were used for GF model and the improved model to simulate respectively. Results were analyzed, which showed that the improved model is much better than the GF model in describing the car-following behavior at the intersections.

关 键 词：交通流 广义力模型 交叉路口 优化速度函数 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]