论文部分内容阅读
针对 LMS算法在噪声主动控制中动态范围小和实时控制跟踪能力不强的缺点 ,提出了多误差改进的归一化变步长最小均方误差自适应滤波算法 ;采用了并行在线辨识算法 (POL I)有效地处理了自适应控制中次级传函识别的关键问题。
Aiming at the shortcoming of LMS algorithm’s small dynamic range and low real-time control tracking ability in active noise control, this paper proposes a modified normalized variable step size minimum mean square error adaptive filtering algorithm based on multiple errors. By using parallel online recognition algorithm (POL I) effectively deal with the key issue of secondary-letter recognition in adaptive control.