A collaborative cuckoo search algorithm with modified operation mode
Qiangda Yang,Huan Huang,Jie Zhang,Hongbo Gao,Peng Liu
DOI: https://doi.org/10.1016/j.engappai.2023.106006
IF: 8
2023-03-01
Engineering Applications of Artificial Intelligence
Abstract:Cuckoo search (CS) is a nature-inspired algorithm that has shown its favorable potential for solving complex optimization problems. Nevertheless, there is a lack of effective information sharing between individuals in CS, which would doubtless limit its achievable performance. While several CS variants have considered this issue, they commonly strengthen the information sharing in just one of the two search parts (i.e., global and local search parts). In this paper, to further address the above issue and to get a more rational allocation of the workloads of global search and local search, a new CS variant called collaborative CS with modified operation mode (CCSMO) is proposed. One novelty is that a collaborative mechanism is presented to strengthen the information sharing and collaboration between individuals in both search parts, and correspondingly, two new iterative strategies are introduced respectively for global search and local search. Another novelty is that the conventional operation mode adopted by almost all existing CS-based algorithms is modified for more rationally allocating the workloads of global search and local search. To validate the performance of CCSMO, extensive experiments and comparisons between CCSMO and 17 state-of-the-art algorithms are made on two popular test suites from IEEE Conference on Evolutionary Computation (CEC). Besides, the algorithm is also applied to solve three engineering design problems and one large-scale combined heat and power economic dispatch problem. The results demonstrate that CCSMO can offer highly competitive performance. Additionally, the time complexity, search behavior, modification effectiveness, and parameter sensitivity of CCSMO are also evaluated.
automation & control systems,computer science, artificial intelligence,engineering, electrical & electronic, multidisciplinary