【作者】 程松； 陈光梦；

【Author】 Cheng Song Chen Guangmeng (Department of Electronic Engineering,Fudan University,Shanghai 200433,China)

【机构】 复旦大学电子工程系；

【摘要】 把自适应滤波器和宏观随机交通流模型结合起来,可以实现对高速公路交通状态的实时估计。高速公路模型被看作是由等距离的路段首尾相接而成的系统,每个路段中交通变量的更新不光与其自身有关,还受到相邻路段的影响。交通传感器设置在路段的交界处,且数量远少于所需估计的状态。本文比较了扩展卡尔曼滤波、无轨迹卡尔曼滤波和粒子滤波三种滤波器在实时交通状态估计问题中的性能,仿真结果表明这三种滤波方法均能够有效地估计和跟踪交通状态的变化,但是在估计精度上粒子滤波器和 UKF 显著地优于扩展卡尔曼滤波器。更多还原

【Abstract】 An approach to the real-time estimation of the traffic state in motorway is developed based on a- daptive filtering and macroscopic stochastic traffic flow model.The motorway stretch is divided into several segments with the same length and the evolution of the traffic variables are influenced by the states of the neighbor segments.Electronic sensors are usually placed between some segments and the measurements are less than states estimated.This paper compares the capabilities of three adaptive filters,which are EKF, UKF and particle filter,in the traffic state estimation issue.Simulation results prove that all of them can predict and track the state efficiently,but the particle filter and UKF is more accurate than the extended Kalman filter.更多还原