基于频率选择性滤波的归一化最小均方算法研究

摘要

Based on the normalized filtered least mean square method which is broadly applied on active noise control technology, the normalized frequency selective FXLMS algorithm and the model for the application of the algorithm on active noise control are calculated and is modeled and analyzed the adaptive attenuation simulation with single tone noise signal and narrow band noise signal demonstrating on the platform of Matlab/Simulink. After the control parameters of filter performance parameters and step size and filter length of adaptive filter is decided. The simulation result is identical with theoretic solution and the presented algorithm is valid.%通过对噪声主动控制技术中广泛应用的归一化滤波最小均方算法的分析,提出了基于频率选择性的归一化滤波最小均方算法,并建立了该算法在噪声主动控制中的应用模型.在Matlab/Simulink环境中建立了归一化频率选择性最小均方算法的仿真模型.通过合理地选取自适应滤波器的结构、滤波器长度、收敛因子等控制参数,对单频噪声信号和窄带随机噪声信号进行了自适应抵消仿真分析.仿真结果表明,噪声信号抵消效果明显,与理论分析的结论一致,证明该算法准确并且有效.