Enhancing Global Optimization through the Integration of Multiverse Optimizer with Opposition-Based Learning

Abstract

The multiverse optimizer (MVO) is increasingly recognized across various scientific disciplines for its robust search capabilities that enhance decision-making in diverse tasks. Despite its strengths, MVO often encounters limitations due to premature convergence, reducing its overall efficiency. To combat this, the study introduces an enhanced version of MVO, termed the improved MVO (iMVO), which incorporates an opposition-based learning (OBL) strategy to overcome this limitation. The effectiveness of iMVO is assessed through a series of tests involving both classical and IEEE CEC 2021 benchmark functions, demonstrating competitive performance against established algorithms. Moreover, the applicability of iMVO to real-world challenges is validated through its successful deployment in civil engineering tasks, particularly in optimizing truss designs and managing time-cost tradeoffs. The results highlight iMVO’s stability and its promising potential for global optimization scenarios.