Competition of Three Chaotic Meta-heuristic Algorithms with Physical Inspiration for Optimal Design of Truss Structures

Abstract

Chaos maps create a significant improvement in the optimization results of meta-heuristic algorithms by creating a balance between the stages of exploration and exploitation. The optimization algorithms of structures are strongly non-linear and non-convex, having several local optima. Chaotic functions, while creating chaotic jumps, provide the conditions for escaping from local optima to global optima. Most of the meta-heuristic algorithms fall into the trap of local optima and suffer some kind of premature convergence. In this paper, by forming three scenarios, chaos functions can be embedded into the exploration, exploitation or both stages at the same time, and improve the results of meta-heuristic algorithms. The considered algorithms are inspired by physical phenomena, with the possibility of accessing classical and regular relations, the effectiveness of chaos functions in meta-heuristic algorithms are increased. Nowadays, chaotic algorithms are widely utilized by researchers and are considered as a challenging topic. In the present research, the effects of logistic and Gaussian chaos functions on the optimization results of three physically inspired meta-heuristic algorithms are investigated. These algorithms include Chaotic Thermal Exchange Optimization (CTEO), Chaotic Big Bang-Big Crunch (CBB-BC), and Chaotic Tug-of-War Optimization (CTWO).