Extended Horizon Model Predictive Control for Cooperative Encirclement of Unmanned Surface Vehicle Swarm

Abstract

Cooperative encirclement of moving targets using unmanned surface vehicle (USV) swarms plays a vital role in maritime patrol, active interception, and autonomous cluster confrontation missions. Conventional finite-horizon model predictive control strategies often suffer from short-sighted control decisions and gradual performance degradation in long-duration cooperative tasks, which may cause unstable formation evolution and potential constraint violations. To address these limitations, this paper proposes an extended horizon model predictive control (EH-MPC) method for USV swarm cooperative encirclement. By extending the optimization prediction horizon, the proposed method endows the controller with long-term decision-making capability, effectively optimizing the swarm formation evolution while strictly maintaining multiple physical constraints, including inter-vehicle safety distance and formation centroid tracking. Comprehensive numerical simulations demonstrate that the EH-MPC strategy can achieve smooth and stable encirclement trajectory evolution, sustained collision-free performance, and high-precision centroid tracking during long-time marine missions. The results verify that the proposed method possesses excellent stability, constraint adaptability, and cooperative consistency for USV swarm encirclement applications.