Publication: Enhanced artificial fish swarm algorithm for optimization
Loading...
Date
2020-08-01
Authors
Sobri, Siti Sarah
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Optimization is an important process in most aspects of daily life, especially in the industrial field. Due to the advancement of technology, the problem of optimization has become more complex and intelligent methods are the best resolution compared to the most conventional methods. Artificial Fish Swarm Intelligent (AFSA) a kind of swarm intelligent algorithm that is widely used in various engineering system designs field and has shown a good performance in solving optimization problem. It has a fast convergence rate and robust to achieve the global optimum because it has the ability to quickly search for a feasible solution in a search space. Despite the advantages, this standard AFSA also faces some problems especially when the complexity of optimization problems increase. Its poor ability to balance between exploration andexploitation has caused slow in convergence rates and difficult to achieve the global optimum. Therefore, this research is aim to improve the convergence rate and global optimization achievement of the AFSA algorithm by introducing a memetic algorithms of Artificial Fish Swarm Algorithm (AFSA). The proposed memetic AFSA is developed through the hybridization of the standard AFSA with a local search technique. Three memetic AFSA algorithms, which are AFSA-HC-F, AFSA-HC-S and AFSA-HC-FS have been proposed to improve the convergence rate and achieve better
global optimum by integrating the Hill Climbing (HC) a local search algorithm into various parts of the standard AFSA. The performance of the standard AFSA and all proposed methods are evaluated based on ten benchmark functions. The results indicated impressive performance in convergence rate and optimum global achievement of the algorithms proposed especially by AFSA-HC-FS compared to the standard AFSA. This shown that these proposed memetic AFSA algorithms have the capability to solve optimization problem better.