To address the problems of search direction imbalance, severe linearization of step size, and high risk of local convergence in traditional swarm intelligence optimization algorithms for high-dimensional complex problems, a novel black bass optimization algorithm based on dynamic information manifolds, curvature coupling dynamics, topological potential field regulation, and entropy-driven energy mechanisms is proposed. The algorithm constructs a dynamic metric space through swarm covariance to update the natural gradient direction; establishes a curvature adjustment mechanism through discrete second-order difference approximation to achieve adaptive burst dynamic regulation; forms an attractive local stable structure by constructing a territory potential field function and a topological deformation tensor; and introduces a swarm information entropy regulation energy system to enable the algorithm to form a non-equilibrium self-regulating mechanism between exploration and exploitation. Theoretical analysis shows that, under the condition that the objective function satisfies Lipschitz continuity, the algorithm's update sequence converges to the neighborhood of the critical point in the desired sense. Complexity analysis shows that the overall time complexity of the algorithm is O(N x D²). This method structurally transforms the search from linear search in Euclidean space to search in a manifold dynamic system.