基于总体平均经验模态分解的主动噪声控制系统研究

发布时间：2018-05-01 13:27

本文选题：混合噪声 + 主动噪声控制　；参考：《自动化学报》2016年09期

【摘要】：为了提高宽窄带混合噪声的消噪效果,本文提出一种基于总体平均经验模态分解(Ensemble empirical mode decomposition,EEMD)的主动噪声控制(Active noise control,ANC)系统,利用实时EEMD算法逐段将混合噪声分解成若干个固有模态函数(Intrinsic mode functions,IMF)分量.因为这些IMF分量的频带各不相同,所以实现了混合噪声中宽带分量和窄带分量的有效分离,独立进行ANC处理后成功解决了处理混合噪声时带来的"火花"现象,而且避免了传统混合ANC(Hybrid ANC,HANC)系统中频率失调的影响.EEMD算法也是对混合噪声的平稳化处理过程,因此当混合噪声中出现非平稳变化时,本文提出的系统也能保持较好的系统稳定性.通过不同噪声环境下进行仿真分析,提出的ANC系统比HANC系统具有更好的系统稳定性和更小的稳态误差.
[Abstract]:In order to improve the denoising effect of wide-band mixed noise, this paper presents an active noise control ANC system based on the total average empirical mode decomposition (EMD). A real-time EEMD algorithm is used to decompose the mixed noise into intrinsic mode functions (IMFs) components. Because the frequency bands of these IMF components are different, the effective separation of wideband component and narrowband component in mixed noise is realized, and the "spark" phenomenon caused by mixed noise processing is solved successfully after independent ANC processing. It also avoids the influence of frequency misalignment in the traditional hybrid ANC(Hybrid ANC system. The algorithm is also a stationary process for the mixed noise. Therefore, the proposed system can maintain good system stability when there is a non-stationary change in the mixed noise. Through simulation analysis in different noise environment, the proposed ANC system has better system stability and smaller steady-state error than HANC system.
【作者单位】：哈尔滨工业大学电气工程及自动化学院;
【基金】：国家自然科学基金(61171183,61471140) 2012航天支撑基金(01320214) 哈尔滨工业大学科研创新基金(HIT.NSRIF201614)资助~~
【分类号】：TB535

【相似文献】