Abstract
A continuous controller is developed for a centralized network control system (NCS) that is composed of agents with nonlinear dynamics subject to a Time-Delay-Switch (TDS) attack and additive disturbances. Since the state tracking error is unmeasurable during TDS attacks, controllers cannot use the state tracking error to coordinate the NCS. Therefore, a novel error signal is designed to address this unique challenge and enable the NCS to achieve a formation control objective. Furthermore, a TDS attack mitigation strategy is developed, which uses both learning and model-based approaches to estimate an agent's state for TDS attack detection and compensation. Lyapunov-Krasovskii (LK) functionals are used in the stability analysis to prove that the tracking errors converge to a steady-state residual, which is a function of the system uncertainty and TDS attack properties. The leader-follower formation control problem for Unmanned Aerial Vehicles (UAVs) based on a pure pursuit guidance law is selected for a simulation study to validate the performance of the proposed method.
Funding Information
  • Air Force Office of Scientific Research (FA9550-19-1-0169)

This publication has 49 references indexed in Scilit: