Re: coprimeunc.m - Hydraulic Manipulator

From: IXP <ixp@leicester.ac.uk>
Date: Fri Sep 13 1996 - 18:31:52 CEST

Dear Ainsley

Sorry about the delay in replying to your questions but I notice Sigurd has
given you some answers. Here is my two-penny-worth:

1. Ill-posed Riccati Equations.

This may not be a problem and often arises. It is warning you that the Riccati
solution may be sensitive. If your solution works then don't worry. In
gamma-iteration the Riccati becomes ill-posed as you approach the optimal
gamma. A sub-optimal gamma is in practice all that is required.

2. If the gamma is too high, then your design is not very robust. You should
therefore reduce your performance demands by reducing the dsired loop gain. You
should then obtain a lower gamma and a better correspondence between the
desired and actual loop shapes.

3. I like to base my design if possible on the best (high order) model and then
reduce the order of the resulting controller.

I hope you find the above comments useful.

Regards

Ian
Received on Fri Sep 13 18:32:09 1996

This archive was generated by hypermail 2.1.8 : Tue Jan 20 2004 - 09:32:14 CET