Quantum Phase-Locked Swordfish Tuna Optimization: A Dissipative Swarm System with Adaptive Subspace Collapse and Irreversible Topological Reconstruction

To address the problems of directional oscillation, uneven energy distribution, dimensional redundancy, and local stagnation in high-dimensional non-convex optimization problems, this paper proposes a swarm intelligent optimization method based on phase manifold dynamics and non-equilibrium dissipative system modeling, termed the Quantum Phase-Locking Skipjack Optimization Algorithm. This method constructs a phase-locked direction field model, representing the individual motion direction as a combination of manifold vectors driven by phase variables, and introduces a Kuramoto-type synchronization mechanism to achieve collaborative evolution of swarm directions. Simultaneously, this paper establishes an open dissipative energy evolution equation to realize an adaptive energy distribution mechanism based on free energy differences. For the high-dimensional space redundancy problem, a variable-dimensional subspace collapse mechanism is proposed. An active dimension set is constructed through a variance threshold, and compression updates are performed within the projected subspace to achieve adaptive manifold dimensionality reduction. Furthermore, an irreversible spike topology reconstruction mechanism is constructed, performing Levy perturbations and topologically repulsive marking on the search path under long-term stagnation conditions. This paper provides a unified model expression from the perspective of continuous-time dynamical systems and constructs a Lyapunov function to analyze system stability. Theoretical analysis shows that the algorithm exhibits directional synchronous convergence and dissipative stability. This model provides a new theoretical framework for the dynamic interpretation and high-dimensional stable optimization of swarm intelligence algorithms.