Research Area:  Metaheuristic Computing

Abstract:

Multi-objective truss optimisation is a research topic that has been less investigated in the literature compared to the single-objective cases. This paper investigates the comparative performance of fourteen new and established multi-objective metaheuristics when solving truss optimisation problems. The optimisers include multi-objective ant lion optimiser, multi-objective dragonfly algorithm, multi-objective grasshopper optimisation algorithm, multi-objective grey wolf optimiser, multi-objective multi-verse optimisation, multi-objective water cycle algorithm, multi-objective Salp swarm algorithm, suc- cess history-based adaptive multi-objective differential evolution, success history–based adaptive multi-objective differential evolution with whale optimisation, non-dominated sorting genetic algorithm II, hybridisation of real-code population-based incremental learning and differential evolution, differential evolution for multi-objective optimisation, multi-objective evolutionary algorithm based on decomposition, and unrestricted population size evolutionary multi-objective optimisation algorithm. The design problem is assigned to minimise structural mass and compliance subject to stress constraints. Eight classical trusses found in the literature are used for setting up the design test problems. Various optimisers are then implemented to tackle the problems. A comprehensive comparative study is given to critically analyse the performance of all algorithms in this problem area. The results provide new insights to the pros and cons of evolutionary multi-objective optimisation algorithms when addressing multiple, often conflicting objective in truss optimisation. The results and findings of this work assist with not only solving truss optimisation problem better but also designing customised algorithms for such problems.

Keywords:  
metaheuristics
truss
optimisation problems
incremental learning
differential evolution

Author(s) Name:  Natee Panagant, Nantiwat Pholdee, Sujin Bureerat, Ali Riza Yildiz, Seyedali Mirjalili

Publisher name:  Springer

DOI:  10.1007/s11831-021-09531-8

Volume Information:  28, pages 4031–4047 (2021)

Paper Link:   https://link.springer.com/article/10.1007/s11831-021-09531-8