Robot Task-Constrained Optimization and Adaptation with Probabilistic Movement Primitives
Enabling a robot to learn skills from a human and adapt to different task scenarios will enable the use of robots in manufacturing to improve efficiency. Movement Primitives (MPs) are prominent tools for encoding skills. This paper investigates how to learn MPs from a small number of human demonstra...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-12-01
|
| Series: | Biomimetics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2313-7673/9/12/738 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1846105611431837696 |
|---|---|
| author | Guanwen Ding Xizhe Zang Xuehe Zhang Changle Li Yanhe Zhu Jie Zhao |
| author_facet | Guanwen Ding Xizhe Zang Xuehe Zhang Changle Li Yanhe Zhu Jie Zhao |
| author_sort | Guanwen Ding |
| collection | DOAJ |
| description | Enabling a robot to learn skills from a human and adapt to different task scenarios will enable the use of robots in manufacturing to improve efficiency. Movement Primitives (MPs) are prominent tools for encoding skills. This paper investigates how to learn MPs from a small number of human demonstrations and adapt to different task constraints, including waypoints, joint limits, virtual walls, and obstacles. Probabilistic Movement Primitives (ProMPs) model movements with distributions, thus providing the robot with additional freedom for task execution. We provide the robot with three modes to move, with only one human demonstration required for each mode. We propose an improved via-point generalization method to generalize smooth trajectories with encoded ProMPs. In addition, we present an effective task-constrained optimization method that incorporates all task constraints analytically into a probabilistic framework. We separate ProMPs as Gaussians at each timestep and minimize Kullback–Leibler (KL) divergence, with a gradient ascent–descent algorithm performed to obtain optimized ProMPs. Given optimized ProMPs, we outline a unified robot movement adaptation method for extending from a single obstacle to multiple obstacles. We validated our approach with a 7-DOF Xarm robot using a series of movement adaptation experiments. |
| format | Article |
| id | doaj-art-069b9d33e37a41d1b5881406bd2fc33e |
| institution | Kabale University |
| issn | 2313-7673 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Biomimetics |
| spelling | doaj-art-069b9d33e37a41d1b5881406bd2fc33e2024-12-27T14:13:25ZengMDPI AGBiomimetics2313-76732024-12-0191273810.3390/biomimetics9120738Robot Task-Constrained Optimization and Adaptation with Probabilistic Movement PrimitivesGuanwen Ding0Xizhe Zang1Xuehe Zhang2Changle Li3Yanhe Zhu4Jie Zhao5State Key Laboratory of Robotics and System, Harbin Institute of Technology, Harbin 150001, ChinaState Key Laboratory of Robotics and System, Harbin Institute of Technology, Harbin 150001, ChinaState Key Laboratory of Robotics and System, Harbin Institute of Technology, Harbin 150001, ChinaState Key Laboratory of Robotics and System, Harbin Institute of Technology, Harbin 150001, ChinaState Key Laboratory of Robotics and System, Harbin Institute of Technology, Harbin 150001, ChinaState Key Laboratory of Robotics and System, Harbin Institute of Technology, Harbin 150001, ChinaEnabling a robot to learn skills from a human and adapt to different task scenarios will enable the use of robots in manufacturing to improve efficiency. Movement Primitives (MPs) are prominent tools for encoding skills. This paper investigates how to learn MPs from a small number of human demonstrations and adapt to different task constraints, including waypoints, joint limits, virtual walls, and obstacles. Probabilistic Movement Primitives (ProMPs) model movements with distributions, thus providing the robot with additional freedom for task execution. We provide the robot with three modes to move, with only one human demonstration required for each mode. We propose an improved via-point generalization method to generalize smooth trajectories with encoded ProMPs. In addition, we present an effective task-constrained optimization method that incorporates all task constraints analytically into a probabilistic framework. We separate ProMPs as Gaussians at each timestep and minimize Kullback–Leibler (KL) divergence, with a gradient ascent–descent algorithm performed to obtain optimized ProMPs. Given optimized ProMPs, we outline a unified robot movement adaptation method for extending from a single obstacle to multiple obstacles. We validated our approach with a 7-DOF Xarm robot using a series of movement adaptation experiments.https://www.mdpi.com/2313-7673/9/12/738human–robot skill transferlearning from demonstrationprobabilistic movement primitivestask-constrained optimizationmovement adaptation |
| spellingShingle | Guanwen Ding Xizhe Zang Xuehe Zhang Changle Li Yanhe Zhu Jie Zhao Robot Task-Constrained Optimization and Adaptation with Probabilistic Movement Primitives Biomimetics human–robot skill transfer learning from demonstration probabilistic movement primitives task-constrained optimization movement adaptation |
| title | Robot Task-Constrained Optimization and Adaptation with Probabilistic Movement Primitives |
| title_full | Robot Task-Constrained Optimization and Adaptation with Probabilistic Movement Primitives |
| title_fullStr | Robot Task-Constrained Optimization and Adaptation with Probabilistic Movement Primitives |
| title_full_unstemmed | Robot Task-Constrained Optimization and Adaptation with Probabilistic Movement Primitives |
| title_short | Robot Task-Constrained Optimization and Adaptation with Probabilistic Movement Primitives |
| title_sort | robot task constrained optimization and adaptation with probabilistic movement primitives |
| topic | human–robot skill transfer learning from demonstration probabilistic movement primitives task-constrained optimization movement adaptation |
| url | https://www.mdpi.com/2313-7673/9/12/738 |
| work_keys_str_mv | AT guanwending robottaskconstrainedoptimizationandadaptationwithprobabilisticmovementprimitives AT xizhezang robottaskconstrainedoptimizationandadaptationwithprobabilisticmovementprimitives AT xuehezhang robottaskconstrainedoptimizationandadaptationwithprobabilisticmovementprimitives AT changleli robottaskconstrainedoptimizationandadaptationwithprobabilisticmovementprimitives AT yanhezhu robottaskconstrainedoptimizationandadaptationwithprobabilisticmovementprimitives AT jiezhao robottaskconstrainedoptimizationandadaptationwithprobabilisticmovementprimitives |