Finite‐Time Containment Control for Multi‐Agent Systems Under DoS Attacks

ABSTRACT This paper studies the finite‐time containment control problem for multi‐agent systems (MASs) with multiple leaders under denial‐of‐service (DoS) attacks that disrupt network connectivity. Unlike traditional consensus problems that drive all agents to a common state, containment control enables follower agents to track and remain within the dynamic convex hull spanned by multiple leaders with different dynamics, offering greater flexibility and applicability in complex multi‐agent coordination tasks. By employing a transformation method, the containment control problem is converted into a stabilization problem. Based on this, a continuous control law is first proposed to ensure that all followers reach the leaders' convex hull in finite time for undirected topologies. To reduce communication load and energy consumption, an event‐triggered control scheme is then developed, guaranteeing convergence to a small neighborhood of the evolving convex hull. For directed topologies, a normalization‐based feedback strategy is introduced to achieve secure finite‐time containment under DoS attacks. Compared with the existing works on containment control, where the topology graph must be undirected or strongly connected, the finite‐time security containment controllers are designed under a directed topology graph and DoS attacks. Numerical simulations verify the effectiveness of the proposed methods.