次级通道在线辨识的齿轮啮合振动主动控制

针对齿轮传动系统中由于啮合误差产生的周期性振动和噪声，构建在从动齿轮轴上附加压电作动器的齿轮主动结构，提出一种次级通道在线辨识的反馈FxLMS算法进行主动控制。应用C-MEX S函数在Simulink中编写了FxLMS算法模块和次级通道进行在线建模的自适应LMS算法模块，仿真算例验证了自建模块的正确性和算法的有效性。将控制算法代码下载到dSPACE中作为控制器，与内置压电作动器的齿轮主动结构组成硬件在环系统进行实验验证。结果表明，在不同啮合频率下，经过主动控制后的齿轮传动系统振动有了不同程度的减弱，在啮合频率基频处有6.9dB的衰减。

Active Vibration Control of Gear Meshing Based on Online Secondary Path Identification Algorithm

Typical gearbox vibration contains several tonal signals mixed with a broadband response. Based on gear active structure attached piezoelectric actuator on the shaft, an online secondary path identification feedback filtered-x LMS (FxLMS) algorithm is proposed to suppress gear meshes vibration due to the transmission error disturbances. Firstly, a C-MEX S function is used to build a Filtered-x LMS algorithm block in Simulink, and then an online secondary path modeling method is added to estimate the secondary path using an adaptive LMS filter. Simulation example verifies the correctness of custom modules and validity of the proposed algorithm. A gearbox test-bed is also developed with a piezoelectric actuator inside for applying control forces to the shaft. Lastly, the custom FxLMS block and the online identification part were downloaded to DSpace as controller. Experiment results showed that the online secondary path modeling FxLMS controller performs an effective performance. Up to 7 dB of gearbox vibration was attenuated at some fundamental gear mesh frequency.