PID controllers used in industry to get a good output response. But tuning the PID controller parameters has to do it manually and it takes lots of effort and timing. Therefore fast tuning techniques required.
Hi Hapuarachchige Don Nelaka Shayamal Priyankara
There are many PID control tuning techniques described in this book that you can explore. By the way, is MathWorks proprietary algorithm for tuning the PID control gains fast enough for your application?
You can relate the PID control parameters to the second order loop feedback system having specific loop gain, damping factor and and natural frequency of oscillations. For optimum system dynamics one uses normally the damping factor to be 0.7. The natural frequency of oscillations is related to the bandwidth of the loop.
In the literature there is already such relations have been found.
Please follow the paper in the link:
PID Tuning Rules for Second Order Systems - Redmine ...
redmine.roboime.com.b
Best wishes
Dear Hapuarachchige Don Nelaka Shayamal Priyankara
I have not built any hardware PID controller in practice yet. However, I have seen in Ogata control engineering textbook, PID control blocks with gains of 0.00001 decimal accuracy. Hence, it seems the gains are to be very accurately tuned. As I know Ziegler Nichols and probably some other PID gain tuning methods, only provide the acceptable values for PID gains to be further tuned probably by software try and error or hardware experiments. Hence, with regard to the required accuracy, I conclude, none of such PID gain tuning methods are reliable in practice, although some actually fumbled tunings will lead to build a PID-controller with a seemingly satisfactory performance. I think for precision control, none of the PID-gain tuning methods are useful. Designing and building an electronic circuitry (including resister and capacitor) with such an accuracy seems to be a tough job.
Noteworthy that, the PID-controller itself has too many deficiencies in its original nature, as it gives divergent control signal when the Derivative gain is added to the controller block, and hence usually PI strategy is used instead of PID, while PI-D, and IPD are two alternative modifications for naïve PID (these are discussed in Ogata control engineering text-book). But these alternatives are also troubled by saturation and slow response time (Ogata). If you want to practically and experimentally design and build a PID-controller, I think it would be only a failed attempt. Many PID-controller designers are accustomed to tune the gains experimentally in the industrial site of the real system (Ogata).
I think only the USA has the capability to design and manufacture such controllers for high precision control tasks, and this is why the USA is the unipolar power in technology.
Please also refer to:
https://www.researchgate.net/post/PID_control_versus_PAD_control_the_Newton_Laplace_principle_of_determinacy
Dear Saeb AmirAhmadi Chomachar
Thank you for your descriptive explanation and reply.
Nelaka
You can use the PID tuner app within Matlab. Also, you can use Optimization Algorithms like PSO, BAT, FPA, and GA, etc through which you can tune the parameters of PID controllers which will give you optimal results.
- Badges
- Science topic
- Similar topics
- Econ
- Manufacturing
- Industry