Projects per year
Abstract
We propose a generalisation of a behaviour tree and motiongenerator based robot arm policy representation for learning and solving tasks such as contact-rich tasks like peg insertion or pushing an object. We use planning to generate skill sequences needed to execute these tasks and rely on reinforcement learning to obtain parameters of the policy. We assume gaussian processes as a suitable method for this generalisation and present preliminary, promising results from initial experiments.
| Original language | English |
|---|---|
| Number of pages | 2 |
| Publication status | Published - 2022 |
Subject classification (UKÄ)
- Robotics
Keywords
- Generalization
- Industrial Robots
- Reinforcement learning
Fingerprint
Dive into the research topics of 'Generalizing Behavior Trees and Motion-Generator (BTMG) Policy Representation for Robotic Tasks Over Scenario Parameters'. Together they form a unique fingerprint.Projects
- 2 Active
-
Efficient Learning of Robot Skills
Mayr, M., Malec, J. & Krueger, V.
2018/09/03 → 2023/09/01
Project: Dissertation
-
RobotLab LTH
Bagge Carlson, F., Johansson, R., Karlsson, M., Olofsson, B., Robertsson, A., Robertz, S., Haage, M., Malec, J., Nilsson, K., Nugues, P., Stenmark, M., Topp, E. A., Krueger, V., Åström, H., Mayr, M., Salt Ducaju, J., Nishimura, M., Wisbrant, J., Dürr, A., Mayr, M., Nugues, P., Klang, M., Klöckner, M., Nardi, L., Ahmad, F., Oxenstierna, J., Rizwan, M., Reichenbach, C., Bergström, J., Dell'Unto, N., Maunsbach, L., Åström, K., Blomdell, A., Soltesz, K., Magnusson, M., Fransson, P., Karayiannidis, Y., Johansson, A. T., Jia, Z., Laban, L., Wingqvist, B., Guberina, M., Jena, A., Westin, E., Frick, C., Pisarevskiy, A., Nilsson, A., Reitmann, S. & Hvarfner, C.
1993/01/01 → …
Project: Research
Equipment
-
RobotLab LTH Infrastructure
Volker Krueger (Manager), Björn Olofsson (Manager), Yiannis Karayiannidis (Manager), Mathias Haage (Manager), Jacek Malec (Manager) & Elin A. Topp (Manager)
Faculty of Engineering, LTHInfrastructure