Grey Mako Shark Optimization Algorithm: A Predator-Inspired Irreversible Momentum Search Framework

To address the problems of premature convergence, dimensional coupling difficulties, and search direction degradation in high-dimensional complex optimization problems, a new optimization algorithm for mackerel sharks based on variable-mass impulse conservation and an information refraction field is proposed. At the dynamic level, the algorithm constructs a state-dependent mass model, introducing impulse conservation and nonlinear dissipation mechanisms to enable adjustable second-order dynamic behavior in individual movements. At the geometric level, an information refraction field is constructed, embedding the local curvature of the objective function into the search update process to adapt the search trajectory to the spatial structure. At the topological level, a predator submanifold locking mechanism is introduced, and adaptive dimensionality reduction search is achieved through dimensional activity analysis. A mirror reflection perturbation mechanism is also designed to suppress stagnation. Theoretical analysis shows that the algorithm constitutes a discrete variable-mass nonlinear dynamic system, which can guarantee stable convergence under certain conditions. The algorithm provides a novel modeling framework for swarm intelligence methods based on the coupling of information geometry and dynamic systems.