Autonomous aerobatic flight has been a challenging control problem for many years. This is because requires such highly precise while operating on the extreme edges of envelope which most controllers are not able to handle. For pilots, this learnt through years experience. The research in paper significantly shortens learning time by extending state art work Deep Reinforcement Learning realm control. presents Normalized Advantage Function controller that, unlike traditional architectures,...

In this paper, we present a Model-Free Stochastic Inverse Optimal Control (IOC) algorithm for the discrete-time infinite-horizon stochastic linear quadratic regulator (LQR). Our proposed exploits richness of available system trajectories to recover control gain <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$K$ </tex-math></inline-formula> and cost function parameters notation="LaTeX">$(Q,R)$ in low...

In this paper, we consider the problem of distributed reachable set computation for multi-agent systems (MASs) interacting over an undirected, stationary graph. A full state-feedback control input such MASs depends no only on current agent's state, but also its neighbors. However, in most MAS applications, dynamics are obscured by individual agents. This makes computation, a fully manner, challenging problem. We utilize ideas polytopic approximation and generalize it to setup. formulate...

Urban air mobility (UAM) systems are characterized by the heterogeneity of participating aerial vehicles (AVs). Participating AVs expected to cooperate with each other while maintaining flexibility in individual missions and reacting possibility cyberattacks security threats. In this paper, we focus on vulnerabilities UAM cyberphysical system against distributed denial-of-service (DDOS) cyberattacks. We develop a resilient control strategy for navigating through airspace mitigate effect DDOS...

The vision for the Urban Air Mobility airspace is a highly automated, cooperative, passenger and/or cargo carrying air transportation service economic purposes. Comprised of complex, safety critical cyber-physical systems (CPSs), integration UAM system within National Airspace System (NAS) requires development robust control paradigms that are resilient to cyberattacks. Consequently, cybersecurity CPSs has emerged as one most important issues general operations. In this paper, we consider...

View Video Presentation: https://doi.org/10.2514/6.2023-2193.vid In this paper, we propose a Semi-Discrete Model-Free Q-Learning control strategy to learn damage recovery policy in sublinear time iterations O( (n + m)/m). We consider the scenario for completely unknown linear time-invariant system, with n- dimensional state space and m-dimensional input space, that has undergone sudden unexpected system damage. setting, allow scarcity on available time, computational resources, rollouts,...

In this paper, we present a provably convergent Model-Free ${Q}$-Learning algorithm that learns stabilizing control policy for an unknown Bilinear System from single online run. Given bilinear system, study the interplay between its equivalent control-affine linear time-varying and time-invariant representations to derive i) Pontryagin's Minimum Principle, pair of point-to-point model-free improvement evaluation laws iteratively solves optimal state-dependent policy; ii) properties under...