Phase portraits as movement primitives for fast humanoid robot control

DOI10.1016/J.NEUNET.2020.04.007zbMATH Open1478.93450DBLPjournals/nn/MaedaKM20arXiv1912.03535OpenAlexW3019662265WikidataQ96160879 ScholiaQ96160879MaRDI QIDQ2055044FDOQ2055044

Authors: Guilherme Maeda, Okan Koç, Jun Morimoto 

Publication date: 3 December 2021

Published in: Neural Networks (Search for Journal in Brave)

Abstract: Currently, usual approaches for fast robot control are largely reliant on solving online optimal control problems. Such methods are known to be computationally intensive and sensitive to model accuracy. On the other hand, animals plan complex motor actions not only fast but seemingly with little effort even on unseen tasks. This natural sense to infer temporal dynamics and coordination motivates us to approach robot control from a motor skill learning perspective to design fast and computationally light controllers that can be learned autonomously by the robot under mild modeling assumptions. This article introduces Phase Portrait Movement Primitives (PPMP), a primitive that predicts dynamics on a low dimensional phase space which in turn is used to govern the high dimensional kinematics of the task. The stark difference with other primitive formulations is a built-in mechanism for phase prediction in the form of coupled oscillators that replaces model-based state estimators such as Kalman filters. The policy is trained by optimizing the parameters of the oscillators whose output is connected to a kinematic distribution in the form of a phase portrait. The drastic reduction in dimensionality allows us to efficiently train and execute PPMPs on a real human-sized, dual-arm humanoid upper body on a task involving 20 degrees-of-freedom. We demonstrate PPMPs in interactions requiring fast reactions times while generating anticipative pose adaptation in both discrete and cyclic tasks.

Full work available at URL: https://arxiv.org/abs/1912.03535

Recommendations

• Adaptive control of humanoid robot movement
• scientific article; zbMATH DE number 1869464
• scientific article; zbMATH DE number 7179809
• A control theory approach to the analysis and synthesis of the experimentally observed motion primitives
• A momentum control for stable whole-body motion of humanoid robots

zbMATH Keywords

reinforcement learningcoupled oscillatorsimitation learningphase estimationmovement primitives

Mathematics Subject Classification ID

Artificial intelligence for robotics (68T40) Automated systems (robots, etc.) in control theory (93C85)

Cites Work

• Model predictive control: Theory and practice - a survey
• Title not available (Why is that?)
• The nature of the coupling between segmental oscillators of the lamprey spinal generator for locomotion: a mathematical model
• A generalized path integral control approach to reinforcement learning
• Nonlinear dynamics and chaos. With applications to physics, biology, chemistry, and engineering
• Dynamical movement primitives: learning attractor models for motor behaviors
• Transfer learning for reinforcement learning domains: a survey
• Reinforcement Learning State Estimator

Cited In (3)

• Title not available (Why is that?)
• Online control of simulated humanoids using particle belief propagation
• Dynamical movement primitives: learning attractor models for motor behaviors

This page was built for publication: Phase portraits as movement primitives for fast humanoid robot control

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2055044)