Design of Adaptive Filter Based on Matlab and Simulink - Advanced Research on Computer Education, Simulation and Modeling - page 292

Design of Adaptive Filter Based on Matlab and Simulink
YanPing He 1 and HaiDong Zhang 2
1 School of Electrician Engineering, Northwest University for Nationalities
2 School of Mathematics and Computer Science, Northwest University for Nationalities,
730030, Lanzhou Gansu, P.R. China
[e-mail], [e-mail]
Abstract. The paper introduces the principle and structure of adaptive filter
based on least mean square algorithm, studies a design scheme of a single fre-
quency adaptive notch filter, and simulates its working procedure by Matlab
programming and Simulink modeling. The simulation results show that the
adaptive notch filters by two methods have filtered noise from noise added signal
in the appropriate parameter.
Keywords: LMS Algorithm, Matlab Simulation, Simulink Simulation, Adaptive
Notch Filter.
1 Introduction
Adaptive filter has an extremely important position in the field of signal processing,
and has been used widely in communication, radar, navigation system and industrial
control, etc. We can not achieve optimal filtering signal with a fixed coefficient filter in
some occasions of unpredictable signal and noise characteristics, and have only a so-
lution that is to introduce adaptive filter [1]. Matlab language has some incomparable
advantages with other high-level languages, such as writing simple, programming with
high efficiency, easy to learn understand, etc. Simulink is one of Matlab toolkits, whose
modeling is the more intuitive and simple than general program modeling. We don't
have to remember usage of all kinds of commands and parameters, and can complete
very complicated work with the mouse [2]. In this paper ,we will carry out quickly and
efficiently adaptive filter analysis, design and simulation in Matlab and Simulink,
change parameters at any time in the design in order to achieve optimization filter, and
observe and compare the filters of two methods design.
2 The Principle of Adaptive Filter Based on LMS Algorithm
The difference of adaptive filter and ordinary filter is that it can dynamically change
parameters varying with outside signal characteristics, and maintain optimal filtering
state. How to adjust parameters according to the change of outside signal is decided by
adaptive algorithm. So adaptive algorithm will influence directly on filtering effect [3].
LMS(Least Mean Square) algorithm is a kind of fast search algorithm of using gradient