Yellowfin Tuna Optimization Algorithm

To address the shortcomings of traditional swarm intelligence optimization algorithms in high-dimensional complex problems, such as insufficient utilization of structural information, premature convergence, isotropic contraction, and diversity degradation, a novel optimization algorithm for yellowfin tuna is proposed. The algorithm abstracts dynamic characteristics from the high-speed cruising, cooperative hunting, and long-distance migration behaviors of yellowfin tuna in the ocean. Mathematically, it constructs a current topological driving field, a covariance manifold compression mechanism, an energy-momentum coupled dynamic system, and a spectral diversity conservation mechanism. A staged contraction mapping design ensures local linear convergence. The algorithm does not rely on the objective function gradient but utilizes the population distribution structure to construct a topological flow field, achieving adaptive structural exploration and anisotropic compression convergence. The paper systematically derives the algorithm's dynamic model and provides convergence and complexity analyses. This method provides a novel framework for swarm intelligence optimization based on distributed topology and manifold dynamics.