I have a question regarding the PI controller. Lets say I have an actuator via which I control some plant. The control signal at the output of the PI controller is at its maximal realizable value. So the actuator supplies 100% power into the plant. Despite this fact the time response of the control loop is sluggish. Does it mean that the actuator has been chosen in wrong manner for a given plant? Thanks for clarification.

  • \$\begingroup\$ If ignoring only the time response (related to I), wouldn't it also be possible that the gain (related to P) was incorrectly chosen for the type of actuator? \$\endgroup\$ – Nedd Mar 24 '19 at 10:23
  • 1
    \$\begingroup\$ @Nedd the control signal is at its maximal value very quickly. So it seems to me that the controller gains are tuned correctly. \$\endgroup\$ – L3sek Mar 24 '19 at 10:30
  • \$\begingroup\$ It may still be possible that the actuator is chosen incorrectly. However, consider for example if P=100, maximal actuator input is 100, and the feedback range is 1 to 100. The actuator will always be at maximum. (Also consider that some systems may require a P <1 or even a -P.) \$\endgroup\$ – Nedd Mar 24 '19 at 11:04
  • \$\begingroup\$ @L3sek, If I understand correctly, you could simply replace your PI controller for a step input and the response would still be slow, correct? Thus the problem has nothing to do with the control loop, but with how your system responds to your actuator. \$\endgroup\$ – xuva Mar 24 '19 at 21:04

If the control signal is at maximum and you are still not getting the system performance you want than the problem would be an under sized actuator.

Consider a controls engineer sitting in their car at a stop light that just turned green. The engineer would like to reach a speed of 40 Mph as soon as possible. So the engineer floors the peddle. You can consider a floored peddle a saturated command signal. At this point it is no longer up to the engineer to affect the acceleration of the vehicle. The only way to improve the cars performance is by changing the car it self. Not the control system (The driver in this case)

| improve this answer | |
  • \$\begingroup\$ thank you for your answer. Lets say I have two actuators. One of them is underrated and the other one is more powerful. I also have a way how to switch from the underrated to the more powerful one. Is there any algorithm for determining the switching instant? For example control error greater than some threshold for some time along with control signal at its maximum value? \$\endgroup\$ – L3sek Mar 25 '19 at 21:17
  • \$\begingroup\$ @L3sek Off hand I do not know. It sounds like you have got a good idea where to start. If this is in regards to a real system would there be a reason to keep both actuators and not simply use the larger one alone? \$\endgroup\$ – Clipboard_Waving_Enginerd Mar 26 '19 at 15:06
  • \$\begingroup\$ I meant for example a battery charger with limited discharging current. This battery charger can be bypassed via parallel contactor. \$\endgroup\$ – L3sek Mar 26 '19 at 20:52

Does it mean that the actuator has been chosen in wrong manner for a given plant?

If the actuator is at a point where it cant give additional energy its called a saturation point. Saturation points are bad because:

  • it makes the system hard to analyze because it no longer behaves linearly.

  • it can make the system unstable because of integrator windup and other effects (the system is not longer linear).

A change to the system would need to be made to prevent saturation of the actuator which could be done by changing the gain or increasing the actuator so it can respond with more energy.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.