A New Hybrid Method Based on Krill Herd and Cuckoo Search for Global Optimisation Tasks

Gai-Ge Wang,Amir H. Gandomi,Xin-She Yang,Amir H. Alavi
DOI: https://doi.org/10.1504/ijbic.2016.10000414
2016-01-01
International Journal of Bio-Inspired Computation
Abstract:Recently, Gandomi and Alavi proposed a new heuristic search method, called krill herd (KH), for solving global optimisation problems. In order to make KH more effective, a hybrid meta-heuristic cuckoo search and krill herd (CSKH) method is proposed for function optimisation. The CSKH introduces krill updating (KU) and krill abandoning (KA) operator originated from cuckoo search (CS) during the process when the krill updating so as to greatly enhance its effectiveness and reliability dealing with numerical optimisation problems. The KU operator inspires the intensive exploitation and allows the krill individuals implement a careful search in the later run phase of the search, while KA operator is used to further enhance the exploration of the CSKH in place of a fraction of the worse krill at the end of each generation. The effectiveness of these improvements is tested by 14 standard benchmarking functions and experimental results show, in most cases, this hybrid meta-heuristic CSKH algorithm is more effective and efficient than the original KH and other approaches.
What problem does this paper attempt to address?