一种用于自适应有源噪声控制的在线次级通道建模方法

摘要

Online secondary path modeling for a feedforward adaptive active noise control system is studied. First of all, the existing online secondary path modeling methods are analyzed and compared and their advantages and disadvantages are given. On this basis, a new online secondary path modeling method based on Fx LMS algorithm is proposed. In this method, the step size of modeling filter is updated by the power of the disturbance and the white noise signal, which can reduce the negative impact of the primary noise on the modeling. Furthermore, the step size of the control filter is updated in accordance with the power of primary noise and the correlation value of error signal. And the random noise effect on the stability of the systems is effectively reduced. Computer simulations show that the proposed method can significantly improve the precision and convergence rate of the secondary path model and reach a good noise reduction effect.%针对前馈式自适应有源降噪系统次级通道在线建模问题，首先分析比较已有的在线次级通道建模方法，给出了各自的优缺点。在此基础上提出了一种基于Fx LMS算法的在线次级通道建模方法，该方法使次级通道建模步长随干扰信号和建模白噪声信号功率自适应更新，降低了初级噪声对建模的不利影响，并利用误差信号相关值和初级噪声能量来更新控制滤波器步长，有效降低了突发随机噪声对系统稳定性的影响。计算机仿真结果表明，该方法显著提高了次级通道模型的精度和系统收敛速度，取得更好的降噪效果。