Improved Learning of Dynamics Models for Control
| Autor: | Lerrel Pinto, Arun Venkatraman, Roberto Capobianco, Daniele Nardi, J. Andrew Bagnell, Martial Hebert |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2017 |
| Předmět: |
Reinforcement learning
Optimal control Dynamics learning Sequential prediction 0209 industrial biotechnology Computer science Control (management) Physical system 02 engineering and technology 010501 environmental sciences Machine learning computer.software_genre 01 natural sciences 020901 industrial engineering & automation 0105 earth and related environmental sciences business.industry System dynamics Dynamics (music) Trajectory Robot Artificial intelligence business computer |
| Zdroj: | Springer Proceedings in Advanced Robotics ISBN: 9783319501147 ISER |
| Popis: | Model-based reinforcement learning (MBRL) plays an important role in developing control strategies for robotic systems. However, when dealing with complex platforms, it is difficult to model systems dynamics with analytic models. While data-driven tools offer an alternative to tackle this problem, collecting data on physical systems is non-trivial. Hence, smart solutions are required to effectively learn dynamics models with small amount of examples. In this paper we present an extension to Data As Demonstrator for handling controlled dynamics in order to improve the multiple-step prediction capabilities of the learned dynamics models. Results show the efficacy of our algorithm in developing LQR, iLQR, and open-loop trajectory-based control strategies on simulated benchmarks as well as physical robot platforms. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|