Improved Asynchronous Krill Herd Algorithm for Global Optimization
Paper ID : 1074-ICTCK
1مصطفی مشکات *, 2مهدی یعقوبی
1Department of Artificial Intelligence, Mashhad Branch, Islamic Azad University, Mashhad, Iran
2کمیته علمی
recently, a novel metaheuristic algorithm named Krill Herd (KH), inspired by the hunting behavior of krills, has been introduced for global optimization. The KH algorithm can perform efficiently during a local search but is unable to escape the local optimum, and hence, cannot achieve an appropriate global search. In this study we improve KH algorithm to solve global optimization problems. For this purpose an asynchronous structure and a probability update position mechanism were used in order to improve the KH search scheme. The proposed algorithm is called improved asynchronous krill herd algorithm (IASKH). IASKH algorithm appears to be well capable than KH for solving various functions. The performance of the method was evaluated on twenty benchmark functions including unimodal, multimodal, and fixed-dimension multimodal benchmark functions. A comparison of the results indicated improved performance of the proposed method as compared to the KH algorithm as well as the most recent optimization methods considered in the present study.
Optimization methods, Global optimization problem, Krill herd algorithm, Quantum-behaved particle swarm optimization
Status : Paper Accepted