基于集员滤波的选择权更新子带有源噪声控制

通过子带自适应滤波结构，可以提高宽带噪声降噪效果，归一化子带自适应滤波（NSAF）通过在每个子带上使用相同的全带自适应滤波器，消除了传统子带结构会在输出端产生混叠分量的问题，具有较好的收敛性能和稳态均方误差。但由于在每个子带上采用相同的全带自适应滤波器，计算量要高于传统子带结构，集员滤波（SMF）技术具有数据选择更新的特点，可有效降低计算复杂度。基于NSAF结构，建立了前馈ANC无延迟结构，并基于集员滤波技术，通过选择部分权更新来进一步减少计算量，仿真验证了该算法对宽带噪声具有更优的降噪效果。

Active noise control with selective partial updates subband architecture based on set-membership filtering

Subband adaptive filter architecture is employed to improve wideband noise control effect. The Normalized Subband Adaptive Filter（NSAF） employs the same fullband adaptive filter in each subband, so that it can eliminate aliasing components in the output of the adaptive filter, having a good convergence performance and steady-state mean square error. However, because of using the fullband adaptive filter in each subband, the computational complexity is higher than conventional subband architec- ture. Set-Membership Filter （SMF） can effectively reduce the computation due to data-selective update. This paper establishes the ANC algorithm based on the NSAF, and combines with SMF and selective partial updates to reduce the computation. A better performance of the algorithm is demonstrated by simulation results.