# Difference between revisions of "HW 7 Problem 2"

m |
m |
||

Line 1: | Line 1: | ||

− | What you're supposed to take away from this problem is a better intuition for how to come up with your 'first cut' controller. By looking at the process transfer function, you can't easily deduce the correct values of the controller gains, but what you should do is look at either the process step response or the process transfer function and reason about | + | What you're supposed to take away from this problem is a better intuition for how to come up with your ''first cut'' controller. By looking at the process transfer function, you can't easily deduce the correct values of the controller gains, but what you should do is look at either the process step response or the process transfer function and reason about |

* how the process needs to be modified to meet the specs | * how the process needs to be modified to meet the specs |

## Revision as of 22:39, 25 November 2007

What you're supposed to take away from this problem is a better intuition for how to come up with your *first cut* controller. By looking at the process transfer function, you can't easily deduce the correct values of the controller gains, but what you should do is look at either the process step response or the process transfer function and reason about

- how the process needs to be modified to meet the specs
- what type of controller (P, PI, PD, PID, etc.) is necessary
- what that controller transfer function should look like

As a side note, if you're convinced that you want a full PID controller, Ziegler-Nichols parameters are a good starting point, but remember that complicated is not always good.

Do NOT simply guess and check gain values. If it is clear that you guessed numbers and they happened to work, you will not get full credit. Explain your thought process in arriving at the controller you choose.

--George Hines 14:38, 25 November 2007 (PST)