Composite Observer-Based Resilient MPC for Heterogeneous UAV-UGV Systems Under Hybrid Cyberattacks

Abstract

This article addresses the model predictive secure control challenge for a heterogeneous system composed of unmanned aerial vehicles (UAVs) and unmanned ground vehicles (UGVs). We aim to cooperatively regulate the UAV-UGV assembly to achieve equilibrium in the presence of external disturbances and hybrid cyberattacks. A composite observer-based resilient model predictive control (CO-RMPC) strategy is introduced. Initially, a composite observer is crafted to estimate system states amidst false data injection attacks, with the estimation confined within a shrinking safety tube by an exponentially tightened constraint integrated into the model predictive control (MPC) framework. In addition, we propose a neighbor output sequence prediction mechanism to mitigate packet loss effects due to denial-of-service attacks on the communication network. We derive sufficient conditions that guarantee the recursive feasibility of the MPC and the stability of the entire closed-loop system. Finally, hardware-in-the-loop (HIL) simulation results confirm the robustness and security efficacy of our CO-RMPC approach.