Newsgroups: comp.ai.neural-nets,sci.cognitive,sci.engr.control
Path: cantaloupe.srv.cs.cmu.edu!nntp.club.cc.cmu.edu!miner.usbm.gov!rsg1.er.usgs.gov!stc06.ctd.ornl.gov!fnnews.fnal.gov!uwm.edu!cs.utexas.edu!news.sprintlink.net!pipex!uknet!festival!udcf.gla.ac.uk!ericr
From: ericr@gryphon.mech.gla.ac.uk (Eric Ronco)
Subject: Modular Neural Adaptive Control System
Message-ID: <D6znts.3Cp@udcf.glasgow.ac.uk>
Sender: news@udcf.glasgow.ac.uk (News)
Organization: University of Glasgow
Date: Thu, 13 Apr 1995 19:37:03 GMT
Lines: 40
Xref: glinda.oz.cs.cmu.edu comp.ai.neural-nets:23430 sci.cognitive:7311 sci.engr.control:5174

dear all,


I am involved in the fram of neurocontrol. More 
specifically, I am conceiving an Adaptive Modular 
Neural Network  algorithm as a controller of 
non linear systems. In a control point of view this 
kind of model leads to a hierachical goal oriented
controler (i.e. some control subgoals has first to be
achieved to enable the whole controler to control
the whole plant).

This is quiet new comparatively to the wildly use approach 
in neurocontrol consisting in mapping the inverse dynamic 
of the system.

An other side of my research, more provocative than effective, 
is to use this autonomous modular system as an system identification
tool. In this case we see modularity as a way to "light"
the inside of the NN "black box".


Any comment on the above purposes will be appreciate.



Any way, I want to apply some ideas on quite simple non linear
control systems as the Inverted Pendulum. Then, I am looking for
some applications. 


So, could you let me know whether you have some
models of control systems which I can used
to test the efficiency of a Modular Neural
Adaptive Control System ?


Thank's in advance

Eric
