Career Advice: How can I prepare for my control engineering graduate studies? 2

[RicoJ]

Hi all,

Your input is REALLY IMPORTANT to me because I am quite confused about where I'm going for my graduate studies in 2 years. Your answer will be an important reference for me and might probably change my life.
Thank you for reading through my questions!

Currently I am a fourth-year electrical engineering student doing a 6-month engineering project management internship at a food company.





1. Which materials to learn before grad school?

I have taken two undergraduate level control system courses. I have learned to tune PID,Q-design, basic MIMO systems and I heard that “Control theory is not so practical and almost died”, while control algorithm (maybe optimization) is promising.Is what I heard true? If so, Should I review some calculus first then look into some optimizations? If so, what optimizations would you recommend?






2. What can I do to get the most out of my internship for control system engineering?

In my job I am working on energy saving projects such as fixing air leaks, installing VFDs, making safety & turn-on/shutoff SOPs, etc. Most of my time I'm just calculating energy savings and look into high efficiency equipment. There is no specific design work involved.

I don’t feel like this job have direct impact on my future control engineering career, even though it does give me great exposure to real machines and how factories work in general. If possible, Could you please help me identify what I can potentially do to get more “direct” experience in control engineering.




Thank you so much for your input!

Rico

 

[PNachtwey]

I have taken two undergraduate level control system courses.

That is a good start. It is not enough and don't think that isall there is.

I have learned to tune PID,Q-design, basic MIMO systems

How did you tune a PID? Trial and error doesn't count.
What is Q-design. The teachers are making up new terms all the time.
MIMO systems aren't basic. I was just at a trade show and the question came up how to control two hydraulic pumps one for the cap end and one for the rod end. Need to control position or force.

and I heard that “Control theory is not so practical and almost died”,

Whoever told you that is a fool. I have made lots of money because I can solve problems others can't. Wait till you get out of university or ask now. Have you EVER seen a piece of equipment with a transfer function attached? I haven't in 35 years. Mechanical people evolve or kludge designs. The only people that seem to really design something are air craft designers where mistake or trial and error is costly. I have seen too many poor designs.

while control algorithm (maybe optimization) is promising.Is what I heard true?

If you mean system identification too then yes? Think of all the transfer functions that were given as problems for you over the years. Where did the transfer function come from?

If so, Should I review some calculus first then look into some optimizations? If so, what optimizations would you recommend?

Your should be able to think in terms of calculus AND differential equations. You should be able to solve a system of differential equations for simulation.
Know BFGS and Levenberg-Marquardt.
Know how to integrate a system of differential equations using Runge-Kutta.
Be able to use state space and Laplace transform for the easy stuff.
Learn pole and zero placement.

I have a YT channel "Peter Ponders PID" where I cover topics in much more depth than your professors. It is meant for control engineers and graduate students and I assume you know the basics. You don't see this stuff in a book. From the analytics I can see that most people bail out on my 20 minute videos after 3 minutes probably because they don't understand the math.

Get a CAS ( computer algebra system ) like Mathematica and Maple. Also learn Python. Download Anaconda.
There is another YT channel I like that shows how to solve control problems in python. The channel is APMonitor. It doesn't go into the control theory as much as I do but it shows how to solve problems quickly in python.





Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

http://forum.deltamotion.com/

 

[RicoJ]

Hi Peter,


Thank you so much for your reply!

Q-design is a also called Youla Parameterization. See

https://en.wikipedia.org/wiki/Youla–Kucera_parametrization

Also I have subscribed to your channel. Thank you!


Rico

 

[lacajun]

Peter, I concur and you have another subscriber. There is always something to learn, when it comes to engineering and control theory especially.

Rico, you will need to brush up on Fourier and Z Transforms for graduate level classes. I would also add filter theory because no process is noiseless. A lot of people let linear algebra escape their minds, once they've graduated, but it is important in controls and in some instrument sectors.

If you can, take a ChemE course to learn mass and energy balance. That's important and used in a lot of industries on a daily basis. Two other courses that will serve you well are fluid dynamics and thermodynamics. Heat transfer is another useful course. Control theory is a broad application of engineering and mathematics.

You'll need to learn about field devices, measurement and control points. They have their own transfer functions and understanding where those are, in the control loop transfer function, is very useful.

Controls is a good field. If you learn instrumentation, controls, and project management, you'll have a good career. I added Statistical Process Control, aka Six Sigma, and other aspects of Total Quality Management to my took box. To me, SPC, Statistical Process Control, is the best fit for engineers that know how to measure and control processes.

The last points I would make are to take the FE during your senior year, get the four years of experience under PEs, and get your PE. Some states will count graduate degrees as one year of experience so check your state's laws.

Pamela K. Quillin, P.E.
Quillin Engineering, LLC
NSPE-CO, Central Chapter
Dinner program:

http://nspe-co.org/events.php