Newsgroups: comp.robotics
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!ix.netcom.com!netcom.com!nagle
From: nagle@netcom.com (John Nagle)
Subject: Re: Biped robot designs & ideas (Question)
Message-ID: <nagleD2IH4M.4zz@netcom.com>
Organization: NETCOM On-line Communication Services (408 261-4700 guest)
References: <3evm76$84n@newsbf02.news.aol.com> <D2CK6w.7pI@armory.com> <yzhu.790268016@saturn.acs.oakland.edu>
Date: Mon, 16 Jan 1995 18:38:46 GMT
Lines: 19

yzhu@saturn.acs.oakland.edu (yzhu) writes:
>Being kicked around are theoretical discussions about static or dynamic
>motion and control. Some of them are quite impressive. But from control
>point of view, by my opinion, there's a thing that cannot be overcome by
>existed theory, which is that the error allowed to foot
>position,velocity and acceleration is zero in some direction, given a
>traditional controller. Now imagine a feedback controller can handle
>only part of its input  and you'll reach similar conclusion as the
>gentleman did as the following.

    Huh?

    Read Raibert.  One, two, and four-legged running machines have been
built.  The control algorithms aren't complicated.  And yes, you can
give his machines a small kick and they recover just fine, even if they
have to hop a few steps backward.

    					John Nagle

