The Collaborative Multi-target Search of Multiple Bionic Robotic Fish Based on Distributed Model Predictive Control

Abstract

In complex water environments, search tasks often involve multiple Autonomous Underwater Vehicles (AUVs), and a single centralized control cannot handle the complexity and computational burden of large-scale systems. Target search in complex water environments has always been a major challenge in the field of underwater robots. To address this problem, this paper proposes a multi-biomimetic robot fish collaborative target search method based on Distributed Model Predictive Control (DMPC). First, we established a bionic robot fish kinematic model and a multi-biomimetic robot fish communication model; second, this paper proposed a distributed model predictive control algorithm based on the distributed search theory framework, so that the bionic robot fish can dynamically adjust their search path according to each other’s position information and search status, avoid repeated coverage or missing areas, and thus improve the search efficiency; third, we conducted simulation experiments based on DMPC, and the results showed that the proposed method has a target search success rate of more than 90% in static targets, dynamic targets, and obstacle environments. Finally, we compared this method with Centralized Model Predictive Control (CMPC) and Random Walk (RW) algorithms. The DMPC approach demonstrates significant advantages, achieving a remarkable target search success rate of 94.17%. These findings comprehensively validate the effectiveness and superiority of the proposed methodology. It can be seen that DMPC can effectively dispatch multiple bionic robot fish to work together to achieve efficient search of vast waters. It can significantly improve the flexibility, scalability, robustness and cooperation efficiency of the system and has broad application prospects.