10.551  Systems Engineering

Congratulations on being done with the class! We've enjoyed having you.

We'd like to hear from you about how the class went, and how we can make it better - please fill out the online class evaluations at http://web.mit.edu/subjectevaluation/. If you've already done so, thanks! If not, please take a couple of minutes to do so before it closes on Monday at 9am - it will really help.

All the best for Practice School, research, and whatever else comes your way in the future!

Jit, on behalf of all of the 10.551 instructors

Hey all,
For both Q1 and Q3, use the corresponding figures in the problem statement to answer the questions. No additional data will be provided.
Best,
Shree

Hey all,
I posted 4 documents under the "projects" tab on Stellar.
1. Problem statement for project 3 on Linear Model Predictive Control
2. A Matlab code that you will need to modify to answer the questions in project 3
3. A paper by Prof. Braatz, which is a good background reading for the project
4. Your groups for project 3.
Best,
Shree

If the interpretation of disturbance 1 of part d in the project statement will cause your team to be unable to submit the report by the original deadline, an extension can be granted. We will accept, without penalty, team submissions for the project on Stellar until 11:59pm, Monday, 30 April (it will be marked LATE, but there will be no penalty if the extension was requested and the report submitted by then). If your team would like to have the extension, please email me BEFORE the original deadline of 8:30am, Thursday, 26th April.

Jit  

A couple of teams raised a question about interpretation of disturbance 1 of part (d) in project 2. Flowrates through a valve are determined by the equation given, which is dependent on delta P across the valve. A change in upstream P will change delta P, in turn changing flowrates for a given stem position. When asked to model such a disturbance, during the split second before any controls come into effect, there will not be a change in the valve stem for the flows of either acetic acid or ethanol. Any controls should not be coded into the disturbance section of the simulation. In addition, the ratio controller also works by adjusting stem values of the ethanol valve in response to acetic acid flow, and not vice versa. There is no feedback control on the acetic acid valve.

Even if it seems that in the model, by adjusting the acetic acid flow, it automatically changes the acetic acid valve position, that does not mean that it is the only way to model the disturbance as described. Could you not model the disturbance more directly by changing Press_In?

Given that it seems like a large portion of the class did not interpret the question in this way, if you have designed a third disturbance that mirrors the above (flow rate changes, not valve stem) interpretation, your solution for disturbance 3 can now be used for disturbance 1, with the appropriate modifications. I will also allow your (different) interpretation for disturbance 1 as your suggested additional disturbance 3 given the short time window remaining, but will require a suggestion of how this could happen in reality.