摘要
为了能更好地实现对车内噪声的控制,提出了一种基于sym6小波的离散小波变换(DWT),将其与FxLMS结合形成DWT-FxLMS主动噪声控制算法,并构建相应的主动噪声控制(ANC)系统模型。将采集的车内噪声作为参考噪声源,参考信号经由MALLAT塔式算法实现的离散小波变换分解为具有多分辨率的多个子带,再由FxLMS算法处理,最终经离散小波逆变换实现信号的重构。利用计算机仿真分析该算法对车内噪声的控制效果并与时域FxLMS算法(TD-FxLMS)和频域FxLMS算法(FD-FxLMS)进行比较。结果表明,与TD-FxLMS算法相比,DWT-FxLMS算法大大降低了计算复杂性且收敛性更好;与FD-FxLMS算法相比,DWT-FxLMS算法能有效消除稳态和非稳态噪声,而FD-FxLMS算法无法有效消除非稳态噪声。
In order to realize a better interior noise control for vehicles, a discrete wavelet transform （DWT） method based on sym6 wavelet is proposed. Combining this DWT method with FxLMS, the DWT-FxLMS active noise control algorithm is formed and the corresponding active noise control （ANC） system model is established. The collected vehicle interior noise is taken as the reference noise source, and the reference signal is decomposed into several sub-bands with multi-resolution through the discrete wavelet transform realized by Mallat pyramidal algorithm. Then, these sub-bands are processed by FxLMS algorithm. Thus, the signal reconstruction is realized by the application of inverse transform of discrete wavelet. The control effect of the proposed algorithm for vehicle interior noise is analyzed by numerical simulation and the results are compared with those of time domain FxLMS algorithm （TD-FxLMS） and frequency domain FxLMS algorithm （FD-FxLMS）. The results show that the DWT-FxLMS algorithm has less computational complexity and better convergence than TD-FxLMS algorithm. Compared with FD-FxLMS algorithm, the DWT-FxLMS algorithm can effectively eliminate stationary and non-stationaxy noises, while FD-FxLMS algorithm cannot effectively eliminate non-stationary noise.
作者
李雅榕
郭辉
王岩松
张亮
刘宁宁
LI Yarong;GUO Hui;WANG Yansong;ZHANG Liang;LIU Ningning(School of Automotive Engineering,Shanghai University of Engineering Science,Shanghai 201620,China)
出处
《噪声与振动控制》
CSCD
2018年第6期65-71,共7页
Noise and Vibration Control
基金
国家自然科学基金资助项目(51675324
51175320)
上海市自然科学基金资助项目(14ZR1418600)
关键词
声学
车内噪声
主动控制
DWT-FxLMS算法
信号重构
多分辨率
acoustics
vehicle interior noise
active control
DWT-FxLMS algorithm
signal reconstruction
multi- resolution
