[ Post a Follow-up ][ New Thread ]

Date: 08/06/99
Posted by: Ramón Scheffer
E-mail: scheffer(at)feq(dot)unicamp(dot)br(no spam)
URL: http://

I am applying an extended kalman filter to a CSTR with a cooling jacket (still simulation study). The temperature and volume are controlled by simple proportional controllers. The filter estimates the states of the CSTR and the controller parameters. I fixed the measurement noise covariance matrix and am adjusting the diagonal elements of the process noise covariance matrix by trial-and-error until satisfactory performance is obtained. Are there methods to obtain the process noise covariance matrix in another way then by trial-and-error giving a better performance???

This message has been here for a period of time, a follow-up seems unnecessary. In this place we have stopped the functions of [post a follow-up]. To discuss and communicate your ideas/info, you may use our new message boards, please go to: Manufacturing Community

When you do want to contact the author of this message, you must replace the "(at)" with "@" and "(dot)" with "." in his/her e-mail address above. But, please do not spam...

To create a new thread, please go to the current board of discussion forums. On that page, click the [Post a Message] or [New Thread] link, then you will find the submit form.

There are more than 55 boards in our site. Please choose, according to your topics, an appropriate one to post your message. If you are new comers, we suggest you to go to our Homepage to have a clear knowledge of the subjects and the orientations of our site before you post any messages.