基于运动学方法和运动几何方法融合的质心侧偏角估计

Vehicle Sideslip Angle Estimation Based on Fusion of Kinematic Method and Kinematic-geometry Method

针对车辆质心侧偏角难以直接测量的问题,提出一种基于运动学方法和运动几何方法融合的质心侧偏角估计方法。基于三自由度车辆运动学模型构建一个扩展卡尔曼滤波器(Extended Kalman filter,EKF),用于质心侧偏角的动态值估计。设计出一种基于车辆运动几何关系的估计方法,利用质心侧偏角与前后轮侧偏角之间的运动几何关系,以及UniTire轮胎模型实现质心侧偏角稳态值的估计。然后根据两种估计方法的特点,设计出一种基于质心侧偏角动态特性的数据融合方法,提高了质心侧偏角估计结果的精度。基于CarSim软件的仿真结果表明,该融合估计方法对传感器信号噪声和误差具有较好的鲁棒性,能够在各种路面条件和驾驶操作条件下实现对车辆质心侧偏角的准确估计。

Considering the vehicle sideslip angle is difficult to measure directly,a fusion estimation method based on kinematic method and kinematic-geometry method is proposed.An extended Kalman filter(EKF)is constructed based on a 3-DoF vehicle kinematics model to estimate the dynamic value of vehicle sideslip angle.And an estimation method based on kinematic-geometry relationship of vehicle is proposed,which utilizes the geometric relationship between vehicle sideslip angle and tire slip angles,as well as UniTire model to estimate the steady-state value of vehicle sideslip angle.Then,according to the applicable domain of the two estimation results,a data fusion method based on the dynamic characteristic of vehicle sideslip angle is designed to improve the accuracy of the final estimation result.The simulation results with CarSim software show that the proposed method is robust to sensor signal noise and errors,and can accurately estimate vehicle sideslip angle under various road conditions and driving maneuvers.