Multi-objective heat transfer search algorithm for truss optimization

• Published in: Engineering with computers 2021-01, Vol.37 (1), p.641-662
• Main Authors: Tejani, Ghanshyam G., Kumar, Sumit, Gandomi, Amir H.
• Format: Article
• Language: English
• Subjects: Algorithms; Benchmarks; CAE) and Design; Calculus of Variations and Optimal Control; Optimization; Classical Mechanics; Computer Science; Computer-Aided Engineering (CAD; Conduction heating; Control; Design optimization; Heat transfer; Heuristic methods; Math. Applications in Chemistry; Mathematical and Computational Engineering; Multiple objective analysis; Optimization; Original Article; Pareto optimization; Rank tests; Search algorithms; Statistical analysis; Systems Theory; Trusses
• ISSN: 0177-0667; 1435-5663
• DOI: 10.1007/s00366-019-00846-6

In the real world, we often come across conditions like optimization of more than one objective functions concurrently which are of conflicting nature and that makes the prospect of the problem more intricate. To overpower this contrasting state, an efficient meta-heuristic (MH) is required, which provides a balanced trade-off between diverging objective functions and gives an optimum set of solutions. In this article, a recently proposed MH called Heat Transfer Search (HTS) algorithm is enforced to elucidate the structural optimization problems with Multi-objective functions (described as MOHTS). MOHTS is an efficient MH which works on the principle of heat transfer and thermodynamics, where search agents are molecules which interact with other molecules and with surrounding through conduction, convection, and radiation modes of heat transfer. Five challenging benchmark problems of truss optimization have been taken into consideration here to examine the effectiveness of MOHTS. Procure results through the proposed method show the predominance over considered MHs. These benchmark problems are considered for discrete design variables for the structural optimization problem with two objectives, namely minimization of truss weight and maximization of nodal displacement. Here, the Pareto-optimal front achieved through computational experiments, in the process of optimization, is evaluated by three distinct performance quality indicators namely the Hypervolume, the Front spacing metric, and Inverted Generational Distance. Also, the obtained results after a number of runs are compared with other existing optimizers in the literature like multi-objective ant system, multi-objective ant colony system, and multi-objective symbiotic organism search, which manifest the superiority in the performance of the proposed algorithm over others. The statistical analysis of the experimental work has been carried out by conducting Friedman’s rank test and Post-Hoc Holm–Sidak test. Graphic abstract