Adaptive consensus control of leader-follower multi-agent system with actuator deception attacks

This work proposes an adaptive distributed cooperative control framework for a multi-agent system (MAS) subjected to unknown disturbance inputs and actuator cyber-intrusions. A resilient model reference adaptive control (MRAC) framework augmented with a distributed disturbance observer (DOB) has been designed and integrated to achieve consensus in a MAS with fast mitigation of unpredictable cyber-threats. The proficiency and useability of the applied control algorithm have been tested on a second-order partially known linear dynamical system encompassing four followers with one leader under undeniable time-varying exogenous disturbance inputs. MAS framework has been set up with a direct communication topology for the leader and an undirected communication topology for follower agents. An in-depth stability analysis of the developed DOB-based MRAC of MAS has been performed and presented using the Lyapunov theorem. The mastery of the designed DOB-based MRAC consensus controller has been quantified by performing extensive comparative studies with reported approaches. The presented results reveal the excellence of the applied cyber-secure control methodology in achieving consensus swiftly among agents with improved steady-state accuracy, active damping of unknown disturbances, and excellent mitigation power.