车辆质心侧偏角估计算法设计与对比分析

Vehicle Sideslip Angle Estimation and Contrastive Analysis

摘要:

车辆质心侧偏角是描述车辆侧向运动状态的重要变量之一，同时也很难准确估计.为此设计了基于运动学的直接积分法，基于扩展卡尔曼滤波的质心侧偏角估计算法和基于广义龙贝格观测器的质心侧偏角估计算法.对信号引入一定强度的噪声，通过多种典型工况对质心侧偏角估计算法进行调试与验证，分析不同算法的特点和工况适应性.直接积分法在长时间尺度下无可用性，但在极限工况下能较为快速地描述质心侧偏角变化趋势；卡尔曼滤波算法整体估计效果较好，但在一些动态工况存在估计偏差，且多为高估；龙贝格观测器算法在大部分工况都能获得良好效果，但在某些情况下存在估计偏差，且多为低估.

Abstract:

It is difficult to estimate the vehicle sideslip angle which is a very important indicator of automotive lateral movement. In order to estimate the sideslip angle accurately, the kinematics direct integral method, the estimation methods based on extended Kalman filter and the extended Luenberger observer were developed in this paper. Then, the three developed sideslip angle estimation methods were adjusted and validated in several typical testing conditions in simulation, with a certain intensity of noise. The characteristics and the adaptability of different methods were also analyzed. The kinematics direct integral method cannot estimate accurately in long time range, but can swiftly show the tendency of the change of sideslip angle in extreme conditions; The extended Kalman filter can generally estimate very well, but it usually overestimates in some dynamic conditions; The extended Luenberger observer can also estimate well, but it often underestimates in some conditions.