Learning control theory first job out of college

[Dextrine]

Hello everyone, so my dilemma is this: I graduated with a degree in physics which had a curriculum that did not cover control theory. I recently got a job as an EE and am now likely to be tasked with designing controllers. How can I most quickly and or thoroughly gain a good applicable understanding of control theory. Book recommendations videos or anything else are very much appreciated. Thank you in advance.
Please excuse brevity and typos. I'm on mobile

 

[Nidum]

Depends just how big the gap in your knowledge is .

Actual control theory is only a small part of the knowledge needed to design control systems .

What sort of controllers are you going to be asked to design ?

 

[JakeBrodskyPE]

You're asking how a professional endeavor works. The answer is very large. It's expensive too. But here it is:

https://www.amazon.com/dp/1466571713/?tag=pfamazon01-20

Don't say I didn't warn you. It's big and it's expensive. And yes, I own a set.

That said, if you're just designing a controller, then reading up on PID control and Ziegler-Nichols tuning methods is a good start. However, there are other methods and concerns that go far beyond these simple methods. I strongly recommend you review Laplace Transforms.

Do note that control theory is a relative of signal theory. The difference is that signal theory is mostly concerned with steady state behavior whereas control theory is mostly concerned with upset response.

Get ready to drink from a fire-hose-- and have fun!

 

[JakeBrodskyPE]

You're asking how a professional endeavor works. The answer is very large. It's expensive too. But here it is:

The Instrument Engineer's Handbook

Don't say I didn't warn you. It's big and it's expensive. And yes, I own a set.

That said, if you're just designing a controller, then reading up on PID control and Ziegler-Nichols tuning methods is a good start. However, there are other methods and concerns that go far beyond these simple methods. I strongly recommend you review Laplace Transforms.

Do note that control theory is a relative of signal theory. The difference is that signal theory is mostly concerned with steady state behavior whereas control theory is mostly concerned with upset response.

Get ready to drink from a fire-hose-- and have fun!

 

[jim hardy]

dang it

twice my post has disappeared

------------------------------------------------------------------------------------------------------

i'd try Schaum's Outline
https://www.amazon.com/dp/0070170525/?tag=pfamazon01-20

here's a brief introduction to controllers
http://www.postreh.com/vmichal/papers/PID-Radio.pdf
Ingenious single amplifier PID is fig 3.

Quick bing search turned up this 158 page pdf...
http://users.auth.gr/kappos/bk-clcon.pdfas Mr Brodsky says - Hang On to your Hat ! Really it's a lot of fun. Mother Nature loves a balance...

 

[Dextrine]

Thank you all so much for the feedback. So far it seems incredibly interesting and I'm sure all your recommendations will add to my intrigue. As for a more specific description of what I'm working in is power electronics. So designing controllers for pwm quite a bit. Any resources on that specifically? I've started reading modern control engineering by ogata which is really great so far

 

[JakeBrodskyPE]

Thank you all so much for the feedback. So far it seems incredibly interesting and I'm sure all your recommendations will add to my intrigue. As for a more specific description of what I'm working in is power electronics. So designing controllers for pwm quite a bit. Any resources on that specifically? I've started reading modern control engineering by ogata which is really great so far

Heheheh, reel him in bit by bit...

This is how it starts. You build a DC motor speed controller with an H bridge and then-- well, what does the motor control? Perhaps it's a propeller for a quad-copter. Well, you're really moving masses of air, and the relationship between the propeller and the mass of air you move to RPM is a knowable thing, so you set up a crude lookup table with interpolation --or you could get cute and do it with a least squares third order polynomial fit. Now you need to adjust thrust and motion of the vehicle. Well, what response are you looking for? What can you measure and where is the dead time in the response? How much noise is in the response?

Before long you're deep into how the device will navigate autonomously and maintain stability for a camera platform.

And then someone wants to scale things up a bit, and now you're looking at building a small generator to run a series of four Variable Frequency Motors...

This is how Control Systems Engineers are brought into the loop...

 

[jim hardy]

Join TI's E2E .

http://e2e.ti.com/support/power_management/
http://www.ti.com/lsds/ti/analog/powermanagement/support-community.page

Train your search engine.

http://www.analog.com/en/products/power-management.html
http://www.intersil.com/en/products/power-management.html

 

[mpresic]

Congratulations on your job. You must have really impressed them to have them hire a physicist to do the same job as an engineer. As of two years ago when I was job-hunting, businesses are looking to fill a selected spot with a very specific skill set that (they believe), they can only fill with an specific engineer with a deep narrow background. For example, they want not just an EE, but an EE with power expertise, or control expertise , or circuit expertise, or ME with propulsions, or ME with structures, etc. Because they had hundreds of openings for a few positions, they could afford to be choosy.

I have a similar story with learning control theory after my physics MS. My employer encouraged and paid for university classes in control theory. In those days my employer actually flew faculty in from across the state to teach classes. They also allowed adjustments in my working hours etc. These were days before computer instruction made long distance instruction less practicable, Now money is tight but your employer may be able to accommodate with tele-classes. If not, things will be harder.

Lot's of smaller less formal classes in programming the Arduino or basic STAMP for beginners. I attended one for Arduino at SPARKFUN in Boulder, Colorado. Perhaps some nearby Community College or local hobby group (check the internet or even the newspaper) can be found.

By the way, I think many professionals with physics degree, enjoy, and are well suited for the transition to control engineering. I sometimes wonder what my life would have been like if I got involved in it before physics.