本論文採用MFC致動器為振動抑制之致動元件,以各種不同之演算法,運算出適當之輸出力,作用於懸臂樑根部,以達到減振之效果。天勾控制為較簡單之速度回授控制方式,實驗的效果相當有限,因此,以主動式控制之FxLMS自適應性演算法,在 控制上有良好的抑振效果,卻因為鑑別所需之權重數量較大,更新時需要採集較多之取樣值,導致抑制振動需要花費較長之穩定時間。對此我們利用陷波濾波器改善此項缺點,並加入 Leaky 演算法,使控制力穩定的收斂而不致於發散。實驗結果證明,該演算法能快速的衰減結構上之振動。 另外,利用CompactRIO完成嵌入式系統之控制方式,搭配I/O模塊擷取振動訊號,使演算法獨立運算於嵌入式系統中。由於振動控制工作於嵌入式系統架構上,取樣頻率可依據需求而大幅提高,並透過Real-Time等功能,穩定且精準的控制系統所需之時間。從實驗的過程和結果可得知,嵌入式系統的確能有效的達到抑振的效果,且控制之穩定度優於電腦端之操作,更適合也更貼近振動控制的運用層面。
This Paper employs a macro-fiber composite (MFC) actuator for vibration suppression in a cantilevered beam due to its higher output force to reject the disturbance. Using a variety of algorithms to calculate the appropriate control action on the root of the cantilever. To achieve the effect of vibration suppression. Skyhook control is relatively simple feedback control, the results have been limited. Therefore, using FxLMS in active adaptive control algorithm has a good effect in vibration suppression control. However, a large number of weights needed to collect more samples to re-identification, it costs a long time to stable. For this drawback, we use notch filter and leaky algorithm. This mothed makes control force stable convergence without divergence. Experimental results show that the algorithm can quickly attenuate vibration of the cantilever.