Toward a theoretical foundation of policy optimization for learning control policies
Gradient-based methods have been widely used for system design and optimization in
diverse application domains. Recently, there has been a renewed interest in studying …
diverse application domains. Recently, there has been a renewed interest in studying …
Global Convergence of Direct Policy Search for State-Feedback Robust Control: A Revisit of Nonsmooth Synthesis with Goldstein Subdifferential
Direct policy search has been widely applied in modern reinforcement learning and
continuous control. However, the theoretical properties of direct policy search on nonsmooth …
continuous control. However, the theoretical properties of direct policy search on nonsmooth …
On the optimization landscape of dynamic output feedback linear quadratic control
The convergence of policy gradient algorithms hinges on the optimization landscape of the
underlying optimal control problem. Theoretical insights into these algorithms can often be …
underlying optimal control problem. Theoretical insights into these algorithms can often be …
Controlgym: Large-scale control environments for benchmarking reinforcement learning algorithms
We introduce controlgym, a library of thirty-six industrial control settings, and ten infinite-
dimensional partial differential equation (PDE)-based control problems. Integrated within the …
dimensional partial differential equation (PDE)-based control problems. Integrated within the …
Global convergence of two-timescale actor-critic for solving linear quadratic regulator
The actor-critic (AC) reinforcement learning algorithms have been the powerhouse behind
many challenging applications. Nevertheless, its convergence is fragile in general. To study …
many challenging applications. Nevertheless, its convergence is fragile in general. To study …
Model-Free -Synthesis: A Nonsmooth Optimization Perspective
In this paper, we revisit model-free policy search on an important robust control benchmark,
namely $\mu $-synthesis. In the general output-feedback setting, there do not exist convex …
namely $\mu $-synthesis. In the general output-feedback setting, there do not exist convex …
A non-iterative approach to linear quadratic static output feedback
HM Escamilla, P Trodden, V Kadirkamanathan - IFAC-PapersOnLine, 2023 - Elsevier
This paper considers the problem of static output feedback (SOF) synthesis for linear time-
invariant (LTI) systems. Static output feedback, and more generally structured controller …
invariant (LTI) systems. Static output feedback, and more generally structured controller …
Two-Timescale Optimization Framework for Decentralized Linear-Quadratic Optimal Control
L Feng, YH Ni, X Zhang - arXiv preprint arXiv:2406.11168, 2024 - arxiv.org
This study investigates a decentralized linear-quadratic optimal control problem, and several
approximate separable constrained optimization problems are formulated for the first time …
approximate separable constrained optimization problems are formulated for the first time …
Accelerated Optimization Landscape of Linear-Quadratic Regulator
L Feng, YH Ni - arXiv preprint arXiv:2307.03590, 2023 - arxiv.org
Linear-quadratic regulator (LQR) is a landmark problem in the field of optimal control, which
is the concern of this paper. Generally, LQR is classified into state-feedback LQR (SLQR) …
is the concern of this paper. Generally, LQR is classified into state-feedback LQR (SLQR) …
Globally Convergent Policy Gradient Methods for Linear Quadratic Control of Partially Observed Systems
While the optimization landscape of policy gradient methods has been recently investigated
for partially observed linear systems in terms of both static output feedback and dynamical …
for partially observed linear systems in terms of both static output feedback and dynamical …