基于噪声补偿的迭代平方根CKF的汽车雷达目标跟踪算法

Target tracking algorithm for automotive radar based on iterated square-root CKF by noise compensation

提出一种适用于汽车雷达目标跟踪的基于噪声补偿的迭代平方根容积卡尔曼滤波(NISRCKF)算法。在继承平方根容积卡尔曼滤波(SRCKF)算法快速和鲁棒等优点的基础上,结合Gauss-Newton迭代理论和噪声补偿方法,设计了一种对SRCKF的量测更新过程进行迭代更新的新算法,充分利用了最新的量测信息,通过选取合适的噪声补偿因子进一步提高对汽车雷达目标跟踪的精度。针对汽车雷达目标跟踪问题进行Monte-Carlo仿真实验,与SRUKF,SRCKF和ISRCKF等经典算法进行对照,实验结果表明,文中算法滤波精度比经典滤波算法有明显的提高,且一定范围内增加迭代次数和设置合适的噪声补偿因子能有效提高滤波精度。

A iterated square-root cubature Kalman filter （ISRCKF） algorithm by the noise compensation for target tracking of automotive radar is proposed. Based on the fast and robust advantages of SRCKF, combined with Gauss-Newton iterative theory and the method for the noise compensation, a new algorithm is designed to iterate and update the measurement process. The newest measurement is fully utilized, and the accuracy of the target tracking can be further improved by selecting the appropriate noise compensa- tion factor. Aimed at target tracking problems to the automotive radar, Monte-Carlo simulation experiments are carried out and the algorithm is compared with classical algorithms, such as square-root unscented Kalman filter （SRUKF）, SRCKF and ISRCKF. The experimental results show that the overall filtering accuracy of the algorithm is obviously improved compared with other classical filtering algorithms. Moreo- ver, increasing the number of iterations in a certain range and setting appropriate noise compensation fac- tor can effectively improve the filtering accuracy.