自适应噪声消除算法的性能比较与仿真

Performance Comparisons and Simulations of Adaptive Noise Cancellation Algorithms

在信号处理中,噪声往往是非平稳和随时间变化的,传统方法很难解决噪声背景中的信号提取问题。通过对自适应噪声消除原理的研究,介绍了基于参考信号和基于预测原理的两种自适应噪声消除(ANC,Adaptive Noise Cancellation)方法,分析对比了基于最小均方(LMS,Least Mean Squares)、递推最小二乘(RLS,Recursive Least Squares)和平方根自适应滤波(QR-RLS,recursive least squares based on QR decomposition)三种噪声消除算法的性能。仿真结果表明:这几种算法都能从高背景噪声中有效地抑制干扰提取出有用信号,显示出了良好的收敛性能。相比之下,RLS算法和QR-RLS算法呈现出更快的收敛速度、更强的稳定性和抑噪能力。

In the signal processing,the noise is often non-smooth and time-varying,so the traditional method is difficult to solve the signal extraction problem from the background noise. Through the study on the principle of adptive noise cancellation, two de-noising method that based on reference signal and principles of prediction have been introduced, and noise canceling performance of the LMS algorithms,RLS algorithms and QR_RLS algorithms were compared. The results of computer simulations show that all of these adaptive algorithms can restrain the disturbance effectively and extract the true signal in strong background noise,shows a good convergence performance. In comparison,the RLS algorithm and QR _RLS algorithm take on faster convergence speed,stronger stability and stronger ability to suppress noise.