基于扩展卡尔曼滤波的轨道垂向不平顺估计

轨道不平顺是影响列车平稳性和舒适度的关键因素,因此及时掌握线路状态对保证列车的运行安全具有重要意义。针对采用单个惯性量较难达到对不同波段不平顺的检测,通过观测多个惯性量,运用扩展卡尔曼滤波解决非线性离散系统的最优估计原理,根据车辆轨道耦合状态空间方程计算递推雅克比矩阵,并结合线性观测方程得到最优状态估计,实现轨道不平顺估计。在Matlab平台下,进行了实测轨道不平顺激励作用下的仿真,将仿真得到的观测值采用本文提出的方法进行轨道垂向不平顺估计,结果表明该算法具有很好的精确性。

Prediction of Vertical Track Irregularities Based on Extended Kalman Filter

Track irregularities are main factors affecting stability and comfort of trains and it is important to understand the status of line to ensure safe operation of trains. Due the difficulties to detect different irregularities of different bands with a single inertial value,the method based on the observation of multi inertia values to predict vertical track irregularity is proposed. The optimal estimation principle of extended Kalman filter for nonlinear discrete systems is used to estimate track irregularities according to Recursive Jacobi matrix,the vehicle track coupling status-space equation and the optimal state estimation obtained with linear measurement equation. On Matlab platform, the simulation of actual track irregularities measured under stimulation is conducted. The simulation results show that this algorithm is accurate to estimate track irregularities.