Neuroadaptive Intrusion‐Tolerant Control for Mixed‐Order Heterogeneous Multiagent Systems: A Fully Actuated System Approach

ABSTRACT

This paper investigates the intrusion‐tolerant control problem for mixed‐order heterogeneous multi‐agent systems (HMASs) with unknown nonlinear dynamics and external disturbances, under deception attacks in the controller‐actuator channel. To address the challenges induced by mixed‐order heterogeneity and directed communication topologies, a hierarchical control architecture is established, incorporating a distributed observer and a decentralized adaptive controller. By transforming the mixed‐order HMASs from state‐space model into a fully actuated system model, the fully actuated system approach (FASA) is integrated with adaptive control to develop the controller, thereby significantly reducing computational complexity compared to traditional backstepping techniques. To counteract the system uncertainties and various types of attacks, a Nussbaum‐function‐based adaptive mechanism is introduced, along with neuroadaptive laws that employ minimal learning parameters. Both theoretical analysis and simulation results demonstrate the effectiveness of the proposed scheme in achieving resilient and adaptive consensus tracking under adversarial conditions.