Safety Formation Based on Control Barrier Function for Multi-agent Systems

A robust containment control strategy for second-order dynamic agent systems is proposed in this paper. Unlike traditional approaches that focus on static or single-target scenarios, the method enables agents to cooperatively surround multiple dynamic targets while ensuring real-time collision avoidance in complex environments. A distributed leader–follower framework is adopted to achieve formation tracking, where only partial agents have access to target information. Control Barrier Functions are employed to enforce safety constraints by guaranteeing forward invariance of the safe set under bounded acceleration. To address potential deadlock issues arising from unreachable ideal formation positions due to environmental obstacles, a mapping-based formation reshaping strategy is proposed. Simulation results in forest-like environments with both static and dynamic obstacles demonstrate that the proposed control strategy successfully achieves target encirclement while maintaining inter-agent and obstacle safety.