This project focuses on the development of robust and resilient control frameworks for autonomous and connected robotic systems operating in complex, uncertain, and adversarial environments. The core objective is to ensure stable, reliable, and high-performance control when systems are subject to communication constraints, external disturbances, model uncertainties, and potential cyber-physical threats. Emphasis is placed on networked autonomous platforms, where reliable coordination and control must be maintained despite limited bandwidth, intermittent connectivity, and malicious disruptions such as Denial of Service (DoS) attacks.
To address these challenges, the project investigates advanced nonlinear and learning-based control strategies, combining resilient control architectures with event-triggered communication mechanisms to reduce network load while preserving system stability. Control designs are grounded in rigorous stability analysis using Lyapunov theory, ensuring provable performance guarantees under uncertainty. Event-triggered and adaptive control structures enable the system to dynamically adjust control actions only when necessary, significantly reducing communication overhead without compromising safety or robustness.
A key application domain of the project is autonomous underwater vehicles (AUVs), which present particularly challenging control problems due to strong nonlinear dynamics, unmodeled hydrodynamic effects, actuator constraints, and limited sensing and communication. The project explores intelligent observer-based control frameworks, integrating neuro-fuzzy reinforcement learning, actor–critic architectures, and high-gain observers to achieve precise trajectory tracking and state estimation in large-scale and unstructured underwater environments. By incorporating adaptive neural and fuzzy models alongside saturation-aware control design, the proposed methods effectively handle unknown dynamics, external disturbances, and actuator limitations, enabling resilient, accurate, and energy-efficient operation of autonomous vehicles in real-world conditions.