Model-based Reinforcement Learning for Optimal Feedback Control of Switched Systems

preview_player
Показать описание
Presented at the 59th Conference on Decision and Control –Jeju Island, Republic of Korea (Dec. 14th-18th, 2020)

This paper examines the use of reinforcement learning-based controllers to approximate multiple value functions of specific classes of subsystems while following a switching sequence. Each subsystem may have varying characteristics, such as different cost or different system dynamics. Stability of the overall switching sequence is proven using Lyapunov-based analysis techniques. Specifically, Lyapunov-based methods are developed to prove boundedness of individual subsystems and to determine a minimum dwell-time condition to ensure stability of the overall switching sequence. Uniformly ultimately bounded regulation of the states, approximation of the value function, and approximation of the optimal control policy is achieved for arbitrary switching sequences provided the minimum dwell-time condition is satisfied.

0:00 Introduction
1:35 Approximate Dynamic Programming (ADP)
3:44 Switched System ADP Problem Formulation
7:37 Assumptions
8:23 Theorem Statements
10:24 Simulations
12:22 Conclusion
  1. UF Nonlinear Controls and Robotics Laboratory
  2. adaptive control
  3. robotics
  4. nonlinear control
  5. UF
  6. University of Florida
Рекомендации по теме
Model-based Reinforcement Learning 0:12:47

Model-based Reinforcement Learning for Optimal Feedback Control of Switched Systems

Model-Based Reinforcement Learning 0:15:14

Model-Based Reinforcement Learning with a Generative Model is Minimax Optimal

Why Choose Model-Based 0:15:01

Why Choose Model-Based Reinforcement Learning?

Model Based Reinforcement 0:27:10

Model Based Reinforcement Learning: Policy Iteration, Value Iteration, and Dynamic Programming

Cooperative Model-Based Reinforcement 0:11:59

Cooperative Model-Based Reinforcement Learning for Approximate Optimal Tracking

Reinforcement Learning Series: 0:21:37

Reinforcement Learning Series: Overview of Methods

DeepRL1.6 Model based 0:10:39

DeepRL1.6 Model based versus Model free Reinforcement Learning Source

Safe Model-based Reinforcement 0:03:00

Safe Model-based Reinforcement Learning with Stability Guarantees (NIPS 2017 Spotlight)

Jeff Dean (Google): 0:43:53

Jeff Dean (Google): Exciting Directions in Machine Learning for Computer Systems at NeurIPS 2024

Some Theoretical Results 0:51:06

Some Theoretical Results on Model-Based Reinforcement Learning by Mengdi Wang

Real-world model-based reinforcement 0:43:39

Real-world model-based reinforcement learning with deep neural networks

The challenges of 1:38:34

The challenges of model-based reinforcement learning and how to overcome them - Csaba Szepesvári

RLSS 2023 - 1:21:42

RLSS 2023 - Model-based Reinforcement Learning - Andreas Krause (presented by Felix Berkenkamp)

[Talk] Planning through 1:01:22

[Talk] Planning through Exploration and Exploitation in Model-based Reinforcement Learning

[HSCC21] M. Cohen 0:23:29

[HSCC21] M. Cohen - Model-Based Reinforcement Learning for Approximate Optimal Control with ...

Model based Reinforcement 0:20:04

Model based Reinforcement Learning

SINDy-RL: Interpretable and 0:21:28

SINDy-RL: Interpretable and Efficient Model-Based Reinforcement Learning

[Paper Summary] Objective 0:05:58

[Paper Summary] Objective Mismatch in Model-based Reinforcement Learning

Efficient model-based reinforcement 0:50:07

Efficient model-based reinforcement learning through optimistic policy search and planning

Model Based RL 0:28:01

Model Based RL Finally Works!

RL + Model-based 0:05:29

RL + Model-based Control: Using On-demand Optimal Control to Learn Versatile Legged Locomotion

Model-Based Reinforcement Learning 0:04:42

Model-Based Reinforcement Learning via Latent-Space Collocation

Learning to Model 0:12:00

Learning to Model What Matters // Model-Based Reinforcement Learning

model based reinforcement 0:00:30

model based reinforcement learning