Analisis transien Least Mean Square filter adaptif

PAGILING, Luther, Prof. Adhi Susanto, M.Sc.,Ph.D

2006 | Tesis | S2 Teknik Elektro

Analisis transien dari filter adaptif diperlukan untuk mengetahui parameter yang mempengaruhi unjuk-kerja sistem, seberapa cepat stabilitas sistem dicapai dan seberapa besar filter adaptif mengurangi noise. Pada Least Mean Square(LMS) adaptif untuk semua nilai varians stabilitas sistem tercapai , nilai varians masukan mempengaruhi mean square deviation (MSD) konvergens. Stabilitas sistem tercapai pada nilai step size (μ), Matriks Korelasi (R), dan banyaknya elemen vektor bobot tertentu. MSD konvergens data normalisasi lebih besar dibandingkan MSD konvergens error linear, demikian halnya MSD konvergens simulasi lebih besar dibandingkan MSD teori. MSD konvergens berbanding lurus dengan nlai varians masukan dan nilai step size.

Transient analysis from adaptive filter required to know about parameters influencing working systems, how fast the reachable system stability and how big adaptive filter can be able to eliminate the noise. At LMS adaptive, all of variants value stability systems reached and assess variant of input influence mean square deviation ( MSD) convergence. Stability of system reached at value of step size, correlation matrix ( R ), and to the number of certain element weight vector. Normalization data convergence by MSD is compared to bigger of linear error convergence from MSD, that way the things of compared to bigger MSD simulation convergence of MSD Theory. MSD Convergence compare diametrical with input variant value and assess step size.

Kata Kunci : Sistem Adaptif, Algoritma LMS