基于两级滤波的车辆相对加速度估计
Estimation of vehicle relative acceleration based on two-level filter

针对智能交通中汽车防追尾碰撞预警技术难以准确获取车辆间相对加速度,从而影响预警准确性的不足,提出了一种基于两级卡尔曼滤波的相对加速度估计方法.首先建立准确描述车辆间相对加速度变化的车辆相对运动模型,利用卡尔曼滤波递推方法初步估计出相对加速度值.然后进一步结合相对速度信息,利用第二级卡尔曼滤波估计相对加速度的误差,对初步估计的相对加速度值进行修正,从而获取较为准确的相对加速度信息.实车试验结果表明,该方法的准确性和可靠性好,加速度估计误差小于0.5 m/s2.
The accurate relative acceleration is difficult to obtain by vehicle rear-end collision warning algorithm in intelligent transportation. In order to meet the accurate requirement of the relative acceleration, an estimation method based on two-level Kalman filter is proposed. The relative motion model of vehicles is established, which can accurately describe the relative acceleration changes. The relative acceleration is preliminary estimated by using Kalman filter. Then the relative speed is employed to estimate the relative acceleration error by using the second level Kalman filter. Therefore, the preliminary estimation results of relative acceleration can be corrected. And an accurate and reliable relative acceleration is accurately and reliablely obtained. The real vehicle test results show that the proposed estimation method has advantages with high accuracy, good reliability and strong adaptability. The acceleration estimation error is less than 0.5 m/s2