Design of Finite Impulse Response Filter Using Chaotic Differential Evolution Algorithm
Loading...
Files
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
In this dissertation work, an approach to design a finite impulse response filter having linear phase
has been proposed. A digital filter design procedure has to obtain such a set of filter coefficients that
expected frequency specification of the signal are met. The proposed approach relies on chaotic
differential evolution evolution algorithm (CDE) to generate the filter coefficients. The chaotic differential
evolution (CDE) algorithm is a modification of conventional differential evolution (DE)
algorithm, which depends on chaotic sequence for its performance improvement. A uniformly
distributed random number generation is very important for better performance of stochastic algorithm.
Chaotic map is similar to uniform random number generating function with better statistical
and dynamic properties. The chaotic sequence provides better balance between the exploration
and exploitation capability of the algorithm. The performance of proposed algorithm is analysed
by experimentation with generalized benchmark problems. Low pass and high pass FIR filters of
20th and 30th order have been designed using CDE approach. Least square error is used as an
objective function to measure the deviation of frequency response of designed filter from the ideal
response. The ripple constraint is introduced in order to reduce the magnitude of ripples present
in pass band and stop band regions of filter. The ripple constraint is taken care of using penalty
method. The simulation results shows that proposed CDE approach with Tent map outperforms
conventional DE algorithm in providing maximum stop band attenuation and minimum stop band
and pass band ripples in frequency response. The application of CDE approach for filtering the
power line interference (PLI) effected electrocardiogram (ECG) signal is also included in this research
work. The experimental results have shown that the proposed approach has ability to design
a filter as per desired specifications.
