Accelerating the Adaptive Algorithm by Fuzzy Theory

• Main Authors: Shin-Yau Chang, 張心垚
• Other Authors: Gwo-Jia Jong
• Format: Others
• Language: zh-TW
• Published: 2003
• Online Access: http://ndltd.ncl.edu.tw/handle/58948194104320744902

碩士 === 國立高雄應用科技大學 === 電子與資訊工程研究所碩士班 === 91 === Adaptive filters are widely employed in various applications such as system identification, channel equalization, signal enhancement, signal prediction and noise cancellation etc. Generally, the Least-Mean-Squares (LMS) adaptive algorithm is used to adaptive filter because of its simplicity and low computation costs, but this algorithm (LMS) still have some disadvantages such as low convergence time and low accuracy. An adaptive algorithm for determining the optimum filer coefficients in an adaptive finite impulse response (FIR) filter is presented. This proposed algorithm, which is the LMS adaptive algorithm combined with the concept of fuzzy logic, to improve the convergence rate of the conventional LMS adaptive algorithm. According to the experiment, these results can be shown the performance of the adaptive filter based on the proposed hybrid algorithm is better than the conventional LMS adaptive algorithm. The proposed algorithm is able to speed up the convergence rate in simulation results. The hybrid algorithm, which is the Recursive Least-Squares (RLS) adaptive algorithm combined with the concept of fuzzy logic, is also shown to apply in adaptive filters.