Looking for an interesting Research Topic


Thread Starter


I am a Firmware engineer with 9 years of professional experience (Embedded Software Engineer, C, Microcontroller, DSP, FPGA etc.) with Control Systems background. I am going to study Control Systems in my MS course.

I am asking from people of industrial automation a good problem for research in the field of Distributed Control Systems and Fieldbus (protocol stack development) which they are facing currently and believe it will benefit them if they have any solution to that specific problem.

My expertise are C,Microcontroller, FPGA, Communication protocol, stack development, Machine Control System design with PLC and intelligent drives. But I am a bit away from the control industry and work more on firmware now.

I would like more if the problem is more of a practical one and it would be interesting if the solution to the problem can generate small scale business after the research even if it may require more research time than MS (I'll consider PHD then).
Self-tuning control loops. The P, I, and D gains adjust themselves. Optimize response to process upsets. I hear it's hard to avoid hunting near the setpoint.

This is quite a hard problem, and perhaps more a math problem, but people would pay good money for this.

I hope you find and interesting topic!

B.O. September 1, 2010

Robert Old
Siemens Industry, Inc.
Building Technologies
1000 Deerfield Pkwy.
Buffalo Grove, IL 60089-4513
bob.old [at] siemens.com

William \(Bill\) L Mostia Jr PE

While I am not sure what the current state of the art is in self-tuners, this is not a new topic as I remember that Foxboro had a self-tuning PID loop turner back in the early '80s based on pattern recognition. There are some self tuners on the market, i.e.

http://www.mstarlabs.com/control/self-tuning-pid.html (experimental)

Here are some papers on the subject.


William (Bill) L. Mostia, Jr. PE
Sr. Consultant
SIS-TECH Solutions, LP

Any information is provided on Caveat Emptor basis.