Formation Tracking for a Multirobotic Fish System via Distributed Model Predictive Control

Fish schooling behavior has attracted significant research interest in recent years due to its biological significance and engineering applications. In this article, a novel distributed nonlinear model predictive control (NMPC) framework is proposed to address the formation tracking control problem of a tail-actuated multirobotic fish system. First, a robotic fish control model is developed by integrating a dynamic model with a central pattern generator (CPG) model, where the oscillations of the robotic fish's caudal fin directly serve as control inputs. Second, dynamic reference trajectories for each robotic fish are generated based on the consensus error, as well as the gait features of neighboring robotic fish and itself, while an adaptive line-of-sight (ALOS) strategy is incorporated into the NMPC framework to enhance tracking performance. Finally, the proposed method is validated through numerical simulations and prototype experiments in two scenarios. The results demonstrate the method's effectiveness, robustness, and scalability, thus offering a promising solution for distributed control in multirobotic fish systems.