Efficient learning of distributed linear-quadratic control policies
From MaRDI portal
Abstract: In this work, we propose a robust approach to design distributed controllers for unknown-but-sparse linear and time-invariant systems. By leveraging modern techniques in distributed controller synthesis and structured linear inverse problems as applied to system identification, we show that near-optimal distributed controllers can be learned with sub-linear sample complexity and computed with near-linear time complexity, both measured with respect to the dimension of the system. In particular, we provide sharp end-to-end guarantees on the stability and the performance of the designed distributed controller and prove that for sparse systems, the number of samples needed to guarantee robust and near optimal performance of the designed controller can be significantly smaller than the dimension of the system. Finally, we show that the proposed optimization problem can be solved to global optimality with near-linear time complexity by iteratively solving a series of small quadratic programs.
Recommendations
- On the sample complexity of the linear quadratic regulator
- Low-complexity learning of linear quadratic regulators from noisy data
- A distributed stochastic approximation algorithm for stochastic LQ control with unknown uncertainty
- Distributed controller design and performance optimization for discrete-time linear systems
- Model-free LQR design by Q-function learning
Cites work
- A Characterization of Convex Problems in Decentralized Control$^ast$
- A convex characterization of distributed control problems in spatially invariant systems with communication constraints
- A System-Level Approach to Controller Synthesis
- Distributed Control of Heterogeneous Systems
- Dynamic system identification. Experiment design and data analysis
- scientific article; zbMATH DE number 708500 (Why is no real title available?)
- scientific article; zbMATH DE number 729680 (Why is no real title available?)
- scientific article; zbMATH DE number 1163807 (Why is no real title available?)
- scientific article; zbMATH DE number 2107836 (Why is no real title available?)
- scientific article; zbMATH DE number 845714 (Why is no real title available?)
- scientific article; zbMATH DE number 961607 (Why is no real title available?)
- Numerical recipes. The art of scientific computing.
- On the complexity of decentralized decision making and detection problems
- On the sample complexity of the linear quadratic regulator
- Separable and Localized System-Level Synthesis for Large-Scale Systems
- Sparse and redundant representations. From theory to applications in signal and image processing.
Cited in
(9)- On the sample complexity of the linear quadratic regulator
- Model-based and model-free designs for an extended continuous-time LQR with exogenous inputs
- scientific article; zbMATH DE number 7625185 (Why is no real title available?)
- Policy Gradient Methods for the Noisy Linear Quadratic Regulator over a Finite Horizon
- Distributed Reinforcement Learning for Decentralized Linear Quadratic Control: A Derivative-Free Policy Optimization Approach
- Linear quadratic control for multiple time-delayed uncertain random systems
- Learning decentralized linear quadratic regulators with \(\sqrt{T}\) regret
- Signed-perturbed sums estimation of ARX systems: exact coverage and strong consistency
- Title not available (Why is no real title available?)
This page was built for publication: Efficient learning of distributed linear-quadratic control policies
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5130900)
