基于平方根容积卡尔曼滤波的路面附着系数估计  

Road Friction Coefficient Estimation Based on Square Root Cubature Kalman Filter

作　　者：杜若飞 王健[1] 李楠 费明哲 邓欢 王云靖 Du Ruofei;Wang Jian;Li Nan;Fei Mingzhe;Deng Huan;Wang Yunjing(Shandong Jiaotong University,Jinan 250357;Ningbo Yinzhou DLT Technologies Co.,Ltd.,Ningbo 315100)

机构地区：[1]山东交通学院,济南250357 [2]宁波市鄞州德来特技术有限公司,宁波315100

出　　处：《汽车工程师》2023年第8期30-37,共8页Automotive Engineer

基　　金：山东省交通运输科技项目(2022B107);2021年山东省高等学校青年创新团队发展计划项目(2021KJ039)。

摘　　要：针对车辆主动安全系统中路面附着系数这一关键参数的估计问题,建立了融合Dugoff轮胎模型的三自由度车辆动力学模型,基于平方根容积卡尔曼滤波(SCKF)思想设计了路面附着系数估计器,该方法将正交三角分解法引入容积卡尔曼滤波(CKF)算法,避免了CKF中对协方差矩阵的开方操作,提高了算法的稳定性。利用MATLAB与CarSim联合仿真平台在不同路面条件下进行了多工况仿真验证,并将所提出算法与CKF算法进行对比,结果表明,SCKF算法的鲁棒性较高,估计精度高于CKF算法的估计精度,可以满足车辆主动安全系统的要求。For the estimation of road adhesion coefficient,a key parameter in vehicle active safety system,a three-degree-of-freedom vehicle dynamics model with Dugoff tire model was established,and the road adhesion coefficient estimator was designed based on Square root Cubature Kalman Filter(SCKF).This method introduced orthogonal trigonometric decomposition into Cubature Kalman Filter(CKF)algorithm,avoiding the root operation of covariance matrix in CKF,and improving the stability of the algorithm.MATLAB and CarSim co-simulation platform were used to carry out multi-condition simulation verification under different road conditions,and the proposed algorithm was compared with CKF algorithm.The results show that SCKF algorithm has high robustness and higher estimation accuracy than CKF algorithm,and can meet the requirements of vehicle active safety system.

关 键 词：路面附着系数估计 平方根容积卡尔曼滤波 车辆动力学模型 Dugoff轮胎模型 

分 类 号：U461.1[机械工程—车辆工程]