Enhance tree-seed algorithm using hierarchy mechanism for constrained optimization problems

• Published in: Expert systems with applications Vol. 209; p. 118311
• Main Authors: Jiang, Jianhua, Meng, Xianqiu, Qian, Lize, Wang, Huan
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 15.12.2022
• Subjects: Continuous optimization problems; Grey Wolf optimizer; MTSA; Tree-seed algorithm; Tree-seed algorithm; MTSA; Grey Wolf optimizer; Continuous optimization problems
• ISSN: 0957-4174; 1873-6793
• DOI: 10.1016/j.eswa.2022.118311

Tree-Seed Algorithm is a kind of swarm intelligence optimization algorithm. It is used to solve various problems widely, but it still has some shortcomings need to be overcame, such as imbalance between exploration and exploitation, local stagnation, premature convergence, and so on. In this study, an enhanced meta-heuristic optimization algorithm, called Migration Tree-Seed Algorithm (MTSA), is proposed inspiring by Grey Wolf Optimizer (GWO). The hierarchical gravity learning and random-based migration mechanisms are introduced to overcome the intrinsic defects of the basic TSA. Firstly, hierarchy mechanism ensures the tree migration to guide the seed generation avoiding the local stagnation. Secondly, random-based migration mechanism increases the seed diversity to improve the exploration ability. Finally, the coordinated update of the two mechanisms actuate a suitable trade-off between exploration ans exploitation. We use IEEE CEC 2014 benchmark functions to compare MTSA with basic TSA, the TSA variants (STSA, EST-TSA, fb_TSA), GWO, ABC, SCA, BOA, JAYA and RSA. MTSA is subsequently applied to three classical engineering design problems reported in the specialized literature. Both results show that the MTSA is very competitive and effective compared with other well-known meta-heuristics, proving its excellent applicability in real-world challenging problems with unknown search spaces. •A modified MTSA is proposed based on the hierarchy mechanism of the GWO and random-based migration mechanism.•Procedure of updating seeds positions allows obtaining neighbouring solutions in terms of α, β, and δ trees for mostly promoting exploitation.•The proposed enhanced MTSA is evaluated using 30 benchmark functions in IEEE CEC 2014 and applying in three engineering problems.•The experimental results were superior to some well-known or variant algorithms in continuous optimization domain.