誤差訊號分離技術用於窄頻噪音主動控制系統之改良

Using Error Signal Separation Techniques to Improve Narrowband Active Noise Control System

陶俊揚(Jun-Yang Tao)

指導教授 ： 張政元

中原大學/工學院/電機工程研究所/碩士(2013年)

https://doi.org/10.6840/cycu201301088

摘要

本論文提出一個新的改良式直接/平行主動噪音控制系統(Active Noise Control, ANC)來進行窄頻噪音的消除，傳統的直接/平行主動噪音控制系統，因為每個濾波器的收斂速度不同，容易造成整個系統的收斂速度及消噪效果降低，而我們分別利用帶阻濾波器與經驗模態分解兩種方式將誤差訊號做分離，每組參考訊號給予相對應的誤差訊號頻率，結合直接/平行主動噪音控制系統做濾波器的權重調整，以提升整個系統的效能。實驗是以MATLAB模擬分析的方式，利用由電腦產生的窄頻噪音訊號和實際擷取到的噪音訊號，分別利用帶阻濾波器及經驗模態分解法在直接/平行形式主動噪音控制系統中來進行，並分析ANC利用直接形式、傳統直接/平行形式搭配兩種誤差訊號分離方法的結果。最後實驗結果證明，搭配帶阻濾波器的直接/平行形式主動噪音控制系統有較好的降噪效果及收斂速度。

關鍵字

直接平行形式 ； 主動式噪音控制系統 ； 帶阻濾波器 ； 經驗模態分解 ； 誤差訊號分離

並列摘要

In this thesis, we proposed a new direct/parallel active noise control (ANC) system to cancel narrowband noise. In general the traditional direct/parallel ANC system has different convergent speed in every channels, worse noise reduction effect and reducing convergent speed of the ANC system. Therefore we use notch filter and empirical mode decomposition (EMD) methods to separate of the signal frequency component, so each adaptive filter uses the separated error signal to update the adaptive filter in the direct/parallel controller with the same frequency component. The simulations by Matlab were done to verify the performance with error signal separation technique. The way by using notch filter provides better performance than EMD.

並列關鍵字

direct/parallel form ； empirical mode decomposition ； EMD ； notch filter ； separate error signal ； active noise system

參考文獻

[2] B. Widrow and S. D. Stearns, “Adaptive Signal Processing,” Englewood Cliffs, NJ:

[3] J. C. Burgess,“Active adaptive sound control in a duct: A computer simulation,”

[4] Denenberg, J.N., "Anti-noise quieting the environment with active noise cancellation technology," Potentials, IEEE , vol.11, no.2, pp.36-40, April 1992

[5] Y. Xiao, L. Ma, K. Khorasano and A. Ikuta, “A new narrowband active noise control system in the presence of sensor error ,” in Proc. 2004 MWSCAS, vol. 2, pp. II-5-II-8, July 2004.

[6] Y. Xiao, A. Ikuta, L. Ma and K. Khorasani, “Stochastic analysis of the FXLMS-based narrowband active noise control system,” IEEE Trans. Audio, Speech and Language Processing, vol. 16, no. 5, pp. 1000-1014, 2008.