基于强跟踪滤波的车辆非线性状态估计

Vehicle nonlinear state estimation based on strong tracking filter

针对车辆稳定性控制过程中较难直接测得的车辆关键状态参数,提出基于强跟踪滤波理论的多传感器线性组合状态最优估计算法.建立包含纵向、侧向及侧倾运动的汽车4自由度非线性动力学模型和状态估计模型,并运用多传感器线性最优融合强跟踪滤波估计器,对汽车关键状态进行仿真分析.分析结果表明,采用该方法可以解决由于模型不确定性造成状态估计值偏离系统真实状态的现象,并能有效抑制滤波发散,具有较大范围的自适应跟踪能力.该方法为汽车先进控制系统中的状态参数估计提供了一种准确且低成本的实时软测量技术.

To solve the problem that some key state parameters in vehicle stability control process are difficult to measure directly,the state optimization estimation algorithm of multi-sensor linear combination based on strong tracking filter is proposed. A four-degree of freedom vehicle nonlinear dynamics model and state estimation model including longitudinal,lateral and roll motion are established. With the estimator of multi-sensor information fusion and the strong tracking filter theory,the vehicle key states are simulated and analysed. The results show that the strong tracking filter offers higher performance potential. It can solve the problem that the state estimation values deviate from the true system states due to the model uncertainty and can inhibit the filtering divergence effectively. The technology of state estimation with the strong tracking filter has a wide range of adaptive tracking capability. It provides a real-time,accurate and low-cost soft-sensor way for vehicle advanced control.