Competition-Driven Equilibrium Optimizer with Boundary Control for Numerical Optimization

Metaheuristic optimization algorithms are widely utilized to solve a broad range of scientific and engineering problems due to their less problem dependence, flexibility, and derivative-free mechanism. Among these, the Equilibrium Optimizer (EO) stands out as a physics-based algorithm grounded in the control volume mass balance model, demonstrating competitive performance against other advanced algorithms. However, the original EO algorithm has some limitations, such as insufficient exploitation capability, a tendency to get stuck in local optima, and an imbalance between exploration and exploitation in complex landscapes. To address these issues, this study introduces an enhanced version of the EO algorithm, called the Competitive-based Equilibrium Optimizer (CBEO). This enhancement includes a novel competition-based learning strategy that establishes a dynamically adaptive mechanism, allowing individuals to compete based on their fitness values. Additionally, a boundary control strategy has been developed to intelligently manage the behavior of individuals at the search space boundaries. The integration of these strategies significantly improves population diversity and global search capability. The exploration and exploitation capabilities of CBEO are further assessed through theoretical analysis and empirical comparisons with 12 other leading optimizers on CEC’ 17 benchmark functions. The experimental and statistical results demonstrate the competitive performance of the proposed algorithm, particularly in terms of solution quality and computational efficiency.