DocumentCode :
2480350
Title :
Adaptive Compensation of a Linearly-Parameterized GMS Friction Model with Parameter Projection
Author :
Nilkhamhang, Itthisek ; Sano, Akira
Author_Institution :
Graduate Sch. of Sci. & Technol., Keio Univ., Yokohama
fYear :
2006
fDate :
13-15 Dec. 2006
Firstpage :
6271
Lastpage :
6276
Abstract :
A new, linearly-parameterized dynamic friction model is proposed based upon the generalized Maxwell-slip (GMS) model. This model is capable of describing essential frictional characteristics, such as the Stribeck effect, hysteresis, stick-slip limit cycling, frictional lag, non-drifting properties, and non-local memory. It replaces the traditional, nonlinear Stribeck function with a new function that is linearly-parameterized, and simplifies the design of an adaptive friction compensator. The proposed compensator has a switching structure to accommodate for the hybrid nature of the GMS model, and the parameter projection method is used to guarantee a bound on parameter estimates and stability of the switching adaptive controller. The validity and effectiveness of the proposed, linearly-parameterized friction compensator is verified by simulations for the velocity control of an inertia system under the influence of dynamic friction
Keywords :
adaptive control; adaptive systems; compensation; friction; parameter estimation; adaptive friction compensator; generalized Maxwell-slip model; inertia system; linearly parameterized dynamic friction model; parameter estimation; parameter projection; switching adaptive controller stability; switching structure; velocity control; Adaptive control; Electrical equipment industry; Friction; Hysteresis; Nonlinear dynamical systems; Parameter estimation; Programmable control; Robust stability; USA Councils; Velocity control;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Decision and Control, 2006 45th IEEE Conference on
Conference_Location :
San Diego, CA
Print_ISBN :
1-4244-0171-2
Type :
conf
DOI :
10.1109/CDC.2006.377695
Filename :
4177845
Link To Document :
بازگشت