半主动悬架的无迹卡尔曼状态观测器设计

卡尔曼滤波器是线性动态系统中应用最广泛的一种状态估计方法。在非线性系统中,扩展卡尔曼滤波(EKF)和无迹卡尔曼滤波(UKF)被广泛应用,相比扩展卡尔曼滤波器,无迹卡尔曼滤波器准确度更高、更易于实现。在车辆动力学这种强的非线性系统中,无迹卡尔曼滤波器应用广泛。设计了一种基于无迹卡尔曼滤波器的半主动悬架系统状态观测器,讨论了不准确的过程噪声协方差Q和测量噪声协方差R、及测量信号组合的选择和不准确的模型参数对状态观测精度的影响,仿真结果表明不准确的过程噪声和测量噪声协方差、不合适的测量信号选择和模型参数不准确的干扰在不同程度上降低了状态估计精度。

State Observer Design for Semi-Active Suspension System Based on Unscented Kalman Filter

The Kalman filter(KF)is probably the most widely used estimation algorithm for dynamic system. However,KF is only reliable for systems that are linear on time scale of the updates. In nonlinear systems,unscented Kalman filter(UKF)is more widely used than extended Kalman filter(EKF)for its higher accuracy and easier implementation in recent years. UKF is also employed in vehicle dynamic system which is serious nonlinear. This paper presents a state observer design for semi-active suspension based on UKF. The effects of incorrect process noise covariance Q and measurement noise covariance R,selection of measurement signals set and parameters disturbance are discussed,and the simulation results show that incorrect noise covariance of process and measurement,unsuitable measurement signals set and model parameter disturbances reduce the estimation accuracy in varying degrees.