A Study of Adaptive Filter Applied to the Infinite Impulse Response System

• Main Authors: Kuei-Teng Cheng, 鄭貴騰
• Other Authors: Yung — Fu Cheng
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/24343785368656266496

碩士 === 國立臺北科技大學 === 電機工程系碩士班 === 92 === This thesis is focused on the performance analysis of adaptive filter applied to the modeling of unknown infinite impulse response (IIR) systems. When an unknown system that is characterized by infinite impulse response is being identified, we can employ a high-order impulse response adaptive filter to approximate it by using equation error method or adaptive delay filter method due to their better performance. Instead of the well-known least mean square (LMS) algorithm, the recursive least square (RLS) algorithm is exploited in the experiment. The experimental results show that optimal nearest solution of each tap-weight in the system can be obtained only after a few times of iteration. Besides, in order to figure out empirical criteria and useful references, the convergence process of learning curve, the way to get delay-taps, and the differences of tap-weights between the unknown system and the proposed adaptive filter are thoroughly investigated and compared.