CaltechAUTHORS
  A Caltech Library Service

Learning Terrain Dynamics: A Gaussian Process Modeling and Optimal Control Adaptation Framework Applied to Robotic Jumping

Chang, Alexander H. and Hubicki, Christian M. and Aguilar, Jeffrey J. and Goldman, Daniel I. and Ames, Aaron D. and Vela, Patricio A. (2020) Learning Terrain Dynamics: A Gaussian Process Modeling and Optimal Control Adaptation Framework Applied to Robotic Jumping. IEEE Transactions on Control Systems Technology . ISSN 1063-6536. (In Press) https://resolver.caltech.edu/CaltechAUTHORS:20200807-102400461

[img] PDF - Accepted Version
See Usage Policy.

2782Kb

Use this Persistent URL to link to this item: https://resolver.caltech.edu/CaltechAUTHORS:20200807-102400461

Abstract

The complex dynamics characterizing deformable terrain presents significant impediments toward the real-world viability of locomotive robotics, particularly for legged machines. We explore vertical, robotic jumping as a model task for legged locomotion on presumed-uncharacterized, nonrigid terrain. By integrating Gaussian process (GP)-based regression and evaluation to estimate ground reaction forces as a function of the state, a 1-D jumper acquires the capability to learn forcing profiles exerted by its environment in tandem with achieving its control objective. The GP-based dynamical model initially assumes a baseline rigid, noncompliant surface. As part of an iterative procedure, the optimizer employing this model generates an optimal control strategy to achieve a target jump height. Experiential data recovered from execution on the true surface model are applied to train the GP, in turn, providing the optimizer a more richly informed dynamical model of the environment. The iterative control-learning procedure was rigorously evaluated in experiment, over different surface types, whereby a robotic hopper was challenged to jump to several different target heights. Each task was achieved within ten attempts, over which the terrain's dynamics were learned. With each iteration, GP predictions of ground forcing became incrementally refined, rapidly matching experimental force measurements. The few-iteration convergence demonstrates a fundamental capacity to both estimate and adapt to unknown terrain dynamics in application-realistic time scales, all with control tools amenable to robotic legged locomotion.


Item Type:Article
Related URLs:
URLURL TypeDescription
https://doi.org/10.1109/tcst.2020.3009636DOIArticle
ORCID:
AuthorORCID
Chang, Alexander H.0000-0001-9036-100X
Hubicki, Christian M.0000-0002-2092-3772
Aguilar, Jeffrey J.0000-0003-3055-3215
Goldman, Daniel I.0000-0002-6954-9857
Ames, Aaron D.0000-0003-0848-3177
Vela, Patricio A.0000-0002-6888-7002
Additional Information:© 2020 IEEE. Manuscript received November 30, 2019; revised March 21, 2020; accepted July 1, 2020. Manuscript received in final form July 13, 2020. This work was supported by NSF under Grant CPS#1544857.
Funders:
Funding AgencyGrant Number
NSFCPS-1544857
Subject Keywords:Gaussian process (GP), learning, optimal control, robotic jumping, terrain dynamics
Record Number:CaltechAUTHORS:20200807-102400461
Persistent URL:https://resolver.caltech.edu/CaltechAUTHORS:20200807-102400461
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:104816
Collection:CaltechAUTHORS
Deposited By: George Porter
Deposited On:10 Aug 2020 15:15
Last Modified:10 Aug 2020 15:15

Repository Staff Only: item control page