基于自适应容积卡尔曼滤波的路面附着系数识别  

Road friction coefficient recognition based on noise adaptive cubature Kalman filter

作　　者：亓佳敖 冯静安[1] 万文康 QI Jiaao;FENG Jing’an;WAN Wenkang(College of Mechanical and Electrical Engineering,Shihezi University,Shihezi,Xinjiang 832003,China;School of Mechanical and Electrical Engineering,Xidian University,Xi’an,Shaanxi 710000,China)

机构地区：[1]石河子大学机械电气工程学院,新疆石河子832003 [2]西安电子科技大学机电工程学院,陕西西安710000

出　　处：《石河子大学学报（自然科学版）》2023年第3期274-278,共5页Journal of Shihezi University(Natural Science)

基　　金：新疆生产建设兵团重大科技项目(2018AA008)。

摘　　要：路面附着系数是车辆行驶稳定性的关键参数之一,精确识别车辆行驶时的路面附着系数是决定车辆安全性能优劣的重要前提。相较通过测量路面物理量的Cause-Based识别方法,基于动力学响应的Effect-Based识别方法受客观环境的影响较小,且经济成本更为节约。本文结合车辆动力学响应与Dugoff轮胎模型公式,基于极大值后验估计(MAP)原理和观测信息对量测噪声的统计特性进行在线估计,并将其嵌入容积卡尔曼(CKF)中构建自适应容积卡尔曼(NACKF)路面附着系数估计器,提高算法的估计精度。CarSim-Simulink仿真试验结果表明,在高附着路面下NACKF算法的估计精度较之传统四维UKF和CKF分别提高了0.001 7和0.000 55,而在对接路面下估计精度较之传统UKF和CKF分别提高了0.172 3和0.039。Road adhesion coefficient is one of the key parameters of vehicle driving stability.Accurate identification of road adhesion coefficient is an important prerequisite for determining vehicle safety performance.Compared with cause-based identification method based on measuring pavement physical quantity,the effect-based identification method based on dynamic response is less affected by objective environment,and the economic cost is more economical.This paper combines vehicle dynamic response with Dugoff tire model formula,based on the principle of maximum a posteriori estimation(MAP) and observation information,the statistical characteristics of measurement noise are estimated online and embedded in CKF(Cubature Kalman Filter) to construct NACKF(Noise Adaptive Cubature Kalman Filter) road adhesion coefficient estimator to improve the estimation accuracy of the algorithm.The results of Carsim-Simulink simulation show that the accuracy of NACKF algorithm is 0.001 7 and 0.000 55 higher than that of traditional four-dimensional UKF and CKF under high adhesion road surface,respectively.Compared with the traditional UKF and CKF,the estimation accuracy is improved by 0.172 3 and 0.039 respectively under the docking road surface.

关 键 词：电动汽车 路面附着系数 自适应容积卡尔曼滤波 极大值后验估计 

分 类 号：U461.91[机械工程—车辆工程]