基于UKF的车辆轨道振动响应优化估计

Optimal estimation for vibration responses of vehicle and track based on UKF

随着高速重载铁路的发展,轮轨之间的相互作用加强,在轨道不平顺激励下车辆轨道的振动响应也随之加剧。提出一种在车辆轨道耦合模型基础上利用UKF(Unscented Kalman Filtering,无迹卡尔曼滤波)算法和车辆测量模型对车辆轨道振动响应进行最优估计的算法。通过在运营车辆的车体、转向架、轴箱上加装加速度传感器、陀螺仪,构建车辆测量模型,利用UKF滤波算法和车辆测量模型对车辆轨道耦合模型输出的轨道振动响应进行非线性滤波,获得最优估计。仿真结果表明,经过UKF滤波算法后的轨道振动响应噪声减小,尤其是轮对加速度、轮轨力和钢轨加速度,明显优于车辆轨道耦合模型直接输出值。

With the development of railway towards higher speed and heavier haul, dynamic interaction be- tween wheel and track is strengthened. Accordingly, the vibration responses of vehicle and track are aggravated un- der the excitation of track irregularity. Based on vehicle track coupling model, an algorithm combining UKF （Un- scented Kalman Filtering） with vehicle measurement model to estimate the vibration responses is proposed in this paper. A measurement model is conceived, which is composed of accelerators on car-body, frame and axle-box and gyroscope on car-body and frame. A nonlinear filtering process using UKF and vehicle measurement model is em- ployed on the output of vehicle track coupling model to achieve the optimal estimation. The simulation results show that the estimated vibration responses of vehicle and track passed through UKF algorithm have less noise, especially for the wheel acceleration, wheel rail force and rail accelerator, which are far better than the direct output of vehicle track coupling model.