基于扩展卡尔曼滤波和希尔伯特-黄变换瞬时频率的齿轮啮合刚度辨识算法

齿轮副是传动系统中的重要部件,齿轮在啮合过程中会出现单、双齿交替参与啮合的情况,造成齿轮啮合刚度周期变化,引起系统振动.齿轮的啮合刚度与齿轮的状态有关,当齿轮出现故障时,齿轮啮合刚度会发生变化,因此通过监测齿轮的啮合刚度就能够估计齿轮副的工作状态.根据齿轮副的动力学模型建立齿轮啮合刚度的离散辨识模型,提出基于扩展卡尔曼滤波器和希尔伯特-黄变换瞬时频率,利用振动信号对齿轮啮合刚度进行估计的动态辨识算法.仿真和实测结果表明,所提出的方法能够跟踪辨识齿轮的啮合刚度,具有较高的辨识精度.

Gear meshing stiffness identification algorithm based on Extended Kalman Filter and HHT instantaneous frequency

Gear pairs were important parts of the transmission system. During the process of meshing, the gear teeth engaged in meshing with single or double numbers alternately which result in periodic change of the meshing stiffness and caused vibration on system. Meshing stiffness related to state of gear pair, failure of gear would change the meshing stiffness. Hence, the working state of gear pairs could be estimated through monitoring the meshing stiffness. The distinct identification model of meshing stiffness of gear pair was established based on dynamic model of gear pair. The vibration signal was used in the identification algorithm of meshing stiffness of gear pairs which was based on the Extended Kalman Filter (EKF) and the instantaneous frequency of Hilbert-Huang transform (HHT). The simulation result and measured result indicate the proposed algorithm can trace the meshing stiffness of gear pair and the identifying accuracy is high.