Abstract:

A model-based estimator of sideslip angle was proposed, and a time-varying parameter was introduced in Magic Formula tire model to be adapted to the roads with different friction coefficients. State space RLS was used and a nonlinear vehicle model was established to design the estimator. The slalom test on the high friction road and three double lance change tests on different friction roads were conducted to evaluate the estimator, and the results show that this method can estimate the sideslip angle accurately even when the vehicle has the large sideslip and the tire takes on the severe nonlinearity. Finally, the RLS is compared with the extended Kalman filter in the terms of the accuracy, computational efficiency and utility condition then to further indicate the performance of the estimator.

Key words: automotive engineering;vehicle sideslip angle;state space model;autoregressive least square(RLS); tire model, nonlinear estimation

摘要：

基于动力学模型对车辆质心侧偏角进行了估计。为使轮胎模型能够适应不同附着系数的路面,将动态参数引入“魔术公式”轮胎侧偏力模型。应用状态空间形式的自回归最小二乘算法（RLS）设计了车辆质心侧偏角估计器。通过车辆在高附着系数路面的蛇形试验和变附着系数路面的双移线试验对估计方法进行了验证，结果表明，即使车辆出现大侧偏情况使轮胎进入到侧偏角—侧偏力特性曲线的非线性区域，提出的估计方法也能够实现对质心侧偏角的估计。将该估计方法与扩展卡尔曼滤波估计在精度、计算效率和使用条件等方面进行了比较，进一步表明所提出方法具有良好性能。

关键词: 车辆工程, 质心侧偏角, 状态空间模型, 自回归最小二乘, 轮胎模型, 非线性估计