高速公路汽车追尾碰撞预警关键参数估计  

Estimation of key parameters of highway rear-end collision warning

作　　者：宋翔[1] 李旭[1] 张为公[1] 

机构地区：[1]东南大学仪器科学与工程学院,江苏南京210096

出　　处：《哈尔滨工程大学学报》2014年第9期1142-1148,共7页Journal of Harbin Engineering University

基　　金：国家自然科学基金资助项目(61273236);教育部博士点基金资助项目(200802861061);江苏省自然科学基金资助项目(BK2010239);江苏省博士后科研基金资助项目(1401012C)

摘　　要：为准确、可靠获取高速公路汽车追尾预警算法的关键参数,提出一种基于车辆相对运动典型工况的估计方法。根据高速公路车辆不同的相对运动典型工况建立多个卡尔曼滤波系统状态模型,以全球定位系统与车车通信信息结合雷达信息作为观测量,并在运行过程中检测、容忍传感器信息的不准确甚至失效,利用交互多模型算法,实时、准确、可靠的获取两车相对距离、速度、加速度以及碰撞时间等关键参数。仿真及实车试验结果表明,估计方法具有精度高、鲁棒性和适应性好的优点,且在传感器失效的情况下依然能取得较好的估计效果。In order to obtain key parameters of highway rear-end warning accurately and reliably, an estimation method is proposed based on typical relative movement working condition of vehicles. According to the different vehicle relative movement conditions on the highway, the multiple Kalman filtering system state models are established. Combining the global positioning system, vehicle-to-vehicle communication information and radar information as observational variables, the key parameters, such as relative distance, relative velocity, relative acceleration and TTC （ time to collision） , are accurately and reliably acquired in real time by using interacting multiple models algorithm. Moreover, the inaccurate and invalid information of sensors is detected and tolerated in this method. The simulation and trial results show that the estimation method has advantages of high accuracy, good robustness and strong adaptability. In addition, the high-quality estimation results can still be obtained even in the case of sensor malfunctions.

关 键 词：追尾碰撞 参数估计 交互多模型 卡尔曼滤波 车车通信 全球定位系统 毫米波雷达 传感器失效容忍 距离碰撞时间 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]