Game theory-based metaheuristics for structural design optimization

摘要

This paper presents game theory-based strategies incorporated with metaheuristics to improve the computation efficiency and effectiveness in design optimization of civil engineering structures. The idea is that a structural optimization problem can be treated as a multiplayer game with optimization algorithms as the players. This study presents an innovative mixed game strategy and provides insights into domain decomposition schemes. Based on game theory strategies, six modified metaheuristic algorithms are proposed and demonstrated through a case study of a 30-story mega-braced frame-core building with a complex configuration. Jaya algorithm is selected as the control to evaluate the proposed methods. The results show that the mixed game strategy helps reduce the computation cost by about 50%, while reducing the weights of the building by 4%, compared with the Jaya algorithm.