Support Vector Machine Networks for Friction Modeling

Research output: Journal Publications and ReviewsRGC 22 - Publication in policy or professional journal

1 Scopus Citations
View graph of relations

Author(s)

Detail(s)

Original languageEnglish
Pages (from-to)2833-2838
Journal / PublicationProceedings of the American Control Conference
Volume4
Publication statusPublished - 2003

Conference

Title2003 American Control Conference (ACC 2003)
LocationThe Adams Mark Hotel
PlaceUnited States
CityDenver
Period4 - 6 June 2003

Abstract

This paper presents a novel model-free parameterization approach of friction modeling for servo-motion systems, where the static friction behavior is parameterized by Support Vector Machine networks. In training such network via SVM regression, the effort of accounting for the complexity variation of the static friction mapping is made in terms of varying smoothness and error-tolerance constraints. It is experimentally demonstrated that the proposed SVM networks can achieve satisfactory friction predictions.