Comparative study on bio-inspired global optimization algorithms in minimal phase digital filters design

Abstract

In this paper, a comparative study is presented on various bio-inspired global optimization algorithms in the problem of digital filters design. The designed digital filters are minimal phase infinite impulse response digital filters with non-standard amplitude characteristics. Due to the non-standard amplitude characteristics, typical filter approximations cannot be used to solve this design problem. In our comparative study, we took into consideration the four most popular bio-inspired global optimization techniques. We examined bio-inspired algorithms such as: an ant colony optimization algorithm for a continuous domain, a particle swarm optimization algorithm, a genetic algorithm and a differential evolution algorithm. After experiments, we observed that the differential evolution algorithm is the most effective one for the problem of the digital filters design. © 2014 Springer International Publishing Switzerland.