A new hybrid chaotic atom search optimization based on tree-seed algorithm and Levy flight for solving optimization problems

Abstract

Optimizing the high computational real-world problems is a challenging task that has taken a great deal of efforts in the last decade. The meta-heuristic algorithms have brought countless benefits. As a result, numerous meta-heuristic algorithms have been developed by getting inspired from natural phenomena. The atom search optimization (ASO) is a physics-based meta-heuristic, which has been developed little while ago. Although ASO is capable of solving various problems, due to low convergence speed and lack of proper balance between exploration and exploitation, it is not efficient enough in sorting out real-world convoluted problems. In the present paper, the convergence speed of ASO is improved using chaotic maps and Levy flight random walk. In addition, ASO is hybridized with the tree-seed algorithm (TSA) to improve exploration and exploitation capabilities and make a proper balance between them. TSA is an innovative intelligent meta-heuristic algorithm that has been inspired by the growth of trees and spreading their seeds and has a decent exploration ability. Furthermore, a novel technique has been applied on the proposed hybrid algorithm as a solution for departure of local optimums. Besides, the effectiveness of our contributions is validated by testing the proposed hybrid algorithm on a vast set of benchmark functions comprising unimodal, multimodal, fixed dimension, shifted–rotated and composite. The obtained results have been compared with several other new and powerful meta-heuristic algorithms in terms of descriptive and inferential statistics. In addition, the algorithms are tested on seven real-life engineering problems. The results of the experiments indicated the effectiveness of contributions and the superiority of the proposed hybrid algorithm over its akin counterparts.