自适应滤波中的许多问题可以用系统识别的观点来解决。建议采用递归的最大似然算法来估计信号模型中的参数。于是用参数估计来形成自适应无限冲激响应滤波器。讨论了若干例子,包括:自适应谱线增强、自适应解卷积、自适应噪对消和自适应延时估计。 Many problems in adaptive filtering can be solved using the system identification point of view. It is recommended to use the recursive maximum likelihood algorithm to estimate the parameters in the signal model. The parameter estimation is then used to form an adaptive infinite impulse response filter. Several examples are discussed, including adaptive spectral line enhancement, adaptive deconvolution, adaptive noise cancellation and adaptive delay estimation.