摘要
伯德灵敏度积分决定反馈有源控制系统具有水床效应,即某个频段内的噪声衰减伴随着另一频段内的噪声放大。已有的反馈系统自适应算法没有明确考虑噪声放大量,非自适应算法考虑了噪声放大量但不能在线调整。为了使自适应反馈系统能明确调节噪声放大量,该文以控制器某个频段内的幅度响应小于指定阈值为约束条件,并作为"惩罚项"与误差信号能量相加得到代价函数,通过最速下降法最小化该代价函数得到了一种频域自适应算法。与已有算法在有源降噪耳机实例中的对比结果表明,该算法能够更为直接地对指定频段内的噪声放大量进行调节,从而在降噪量与噪声放大量之间取得折衷。
The waterbed effect of the feedback active noise control system is determined by the Bode's sensitivity integral,which states that if the noise is attenuated at some frequencies,there is always some noise amplification at some other frequencies.The offline design methods for the feedback active noise control system take the noise amplification into account,but they are not adaptive,the adaptive systems do not consider the waterbed effect explicitly.A frequency domain filtered-x least mean square algorithm is introduced to tune the noise amplification in this paper where a constraint that limits the gains of the controller filter at some discrete frequencies are defined and merged into the error signal energy as a penalty term to obtain the new cost function,the update equation of the controller filter is given by minimizing the new cost function with the steepest descent method.Simulations in an active noise control headphone application show that the proposed algorithm can tune the noise amplification more effectively than the existing algorithms.
引文
[1]张振超,安峰岩,吴鸣,等.一种数字有源降噪耳机控制器设计方法[J].应用声学,2017,36(4):317-323.Zhang Zhenchao,An Fengyan,Wu Ming,et al.A method for controller design of digital active noise control headsets in China[J].Journal of Applied Acoustics,2017,36(4):317-323.
[2]郑洋,唐加能.应用于声反馈抑制的变步长归一化子带自适应滤波算法[J].应用声学,2018,37(3):356-364.Zheng Yang,Tang Jianeng.A variable step size normalized subband adaptive filters for acoustic feedback cancellation in China[J].Journal of Applied Acoustics,2018,37(3):356-364.
[3]武帅兵,吴鸣,杨军.参量阵扬声器在管道噪声控制中的研究[J].应用声学,2013,32(6):439-445.Wu Shuaibing,Wu Ming,Yang Jun.An investigation of parametric array loudspeakers in active noise control of a duct in China[J].Journal of Applied Acoustics,2013,32(6):439-445.
[4]Elliott S J.Signal processing for active control[M].London:Academic Press,2001:5-10.
[5]陈克安.有源噪声控制[M].北京:国防工业出版社,2014:4-12.
[6]Bai M,Lee D.Implementation of an active headset by using the H∞robust control theory[J].Journal of the Acoustical Society of America,1997,102(4):2184-2190.
[7]Rafaely B,Elliott S J.H2/H∞active control of sound in a headrest:design and implementation[J].IEEE Transactions on Control Systems Technology,1999,7(1):79-84.
[8]Zhang L,Tao J,Qiu X.Active control of transformer noise with an internally synthesized reference signal[J].Journal of Sound and Vibration,2012,331(15):3466-3475.
[9]斯科格斯特德S,波斯尔思韦特I.多变量反馈控制--分析与设计[M].韩崇昭,张爱民,刘晓风,译.西安:西安交通大学出版社,2011:144-145.
[10]张丽敏,祖峰磊,邱小军.反馈有源噪声控制系统水床效应的分析与改善[J].应用声学,2010,29(6):401-408.Zhang Limin,Zu Fenglei,Qiu Xiaojun.Analysis and improvement on waterbed effect of feedback active noise control system in China[J].Journal of Applied Acoustics,2010,29(6):401-408.
[11]Luo L,Sun J,Huang B.A novel feedback active noise control for broadband chaotic noise and random noise[J].Applied Acoustics,2017,116:229-237.
[12]Wu L,Qiu X,Guo Y.A simplified adaptive feedback active noise control system[J].Applied Acoustics,2014,81(6):40-46.
[13]Wu L,Qiu X,Burnett I S,et al.Decoupling feedforward and feedback structures in hybrid active noise control systems for uncorrelated narrowband disturbances[J].Journal of Sound and Vibration,2015,350:1-10.
[14]Shynk J J.Frequency-domain and multirate adaptive filtering[J].IEEE Signal Processing Magazine,2002,9(1):14-37.
[15]毛鑫,卢晶,邹海山.频域自适应算法在有源噪声控制系统中的性能研究[J].南京大学学报:自然科学,2014,50(1):41-45.Mao Xin,Lu Jing,Zou Haishan.Investigation of the performance of frequency domain adaptive filtering algorithms in active noise control system in China[J].Journal of Nanjing University,2014,50(1):41-45.
[16]Tobias O J,Seara R.Leaky-FxLMS algorithm:stochastic analysis for Gaussian data and secondary path modeling error[J].IEEE Transactions on Speech and Audio Processing,2005,13(6):1217-1230.
