Newsgroups: comp.robotics
Path: brunix!uunet!usc!sol.ctr.columbia.edu!destroyer!ubc-cs!fornax!nitish
From: nitish@cs.sfu.ca (Nitish Mandal)
Subject: Cartesian Model-Based controllers
Message-ID: <1992Aug16.001819.16508@cs.sfu.ca>
Organization: CSS, Simon Fraser University, Burnaby, B.C., Canada
Date: Sun, 16 Aug 1992 00:18:19 GMT
Lines: 22

Hi, I am working on a 2DOF DD robot. We have a DSP controller,
which permits us very low sampling times. Recently we implemented
a joint space model-based controller, and the performance is quite
good. My thesis is in the area of force control. From my experience with
model-based schemes, I have found sampling time to be crucial to their
performance, especially since we plan to use it for force control. My 
question is, has anyone implemented a cartesian-model-based scheme on 
a robot? If so, how was the performance? What is the approximate
increase in computational time? Is it really of any advantage to
use a cartesian scheme, especially for force control?

Any help would be useful.


Thanks,
-Nitish.
nitish@cs.sfu.ca





