Deep-Sea Grouper Optimization Algorithm

To address the common challenges of nonlinearity, multimodality, and susceptibility to local optima in complex continuous optimization problems, this paper proposes a novel swarm intelligence optimization method—the Deep-Sea Grouper Optimization Algorithm (DSGOA). This algorithm uses the lurking, cruising, cooperative hunting, and deep migration behaviors of deep-sea groupers under high-pressure environments as a biological prototype, constructing a new optimization mechanism from three levels: search dynamics, spatial structure, and individual state scheduling. First, the objective function value is mapped to virtual deep-sea pressure, introducing a pressure gradient as the search driving force to achieve directional search under gradient-free conditions. Second, by constructing a multi-depth search space, the group adaptively migrates between different levels, thus forming a dynamic balance between global exploration and local exploitation. Simultaneously, a lurking memory pool is introduced to retain historical potential solutions. Combined with an asymmetric cooperative hunting mechanism and an energy-pressure coupled scheduling model, the algorithm's ability to escape local optima and its search stability are significantly improved. The algorithm has a clear structure, well-defined physical meaning of its parameters, good theoretical interpretability, and broad application prospects.