home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.engr.control
- Path: sparky!uunet!spool.mu.edu!sol.ctr.columbia.edu!destroyer!ubc-cs!unixg.ubc.ca!kakwa.ucs.ualberta.ca!vm.ucs.UAlberta.CA!PH19
- From: PH19@vm.ucs.UAlberta.CA
- Subject: Parameterizing control systems
- Message-ID: <1992Sep13.213120.12709@kakwa.ucs.ualberta.ca>
- Sender: news@kakwa.ucs.ualberta.ca
- Nntp-Posting-Host: vm.ucs.ualberta.ca
- Organization: University Of Alberta, Edmonton
- Date: Sun, 13 Sep 1992 21:08:00 GMT
- Lines: 18
-
- Hi;
-
- I'm trying to estimate the parameters of a fairly simple classical control
- system. There are n parallel systems each of which is a simple second
- order system using the same paramters (ie, all n systems have identical
- parameters). The inputs to these systems are known, but the outputs
- are not, rather what I have is a stream of data showing the identity of
- the system that had the highest output at that time. From this data
- I hope to find out the parameters. Has anyone any suggestions as to
- where I might look to find out how to do it? Or better yet, has anyone
- any idea of how it might be done. Oh, and for the curious, this is a
- model of a consumer choice process where the data is the purchased
- brand and timing and the inputs are the pertinent marketing variables
- for each of n products.
-
- Any help would be greatly appreciated.
-
- Chris Dubelaar
-