You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is your feature request related to a problem? Please describe.
When actuator saturates, the integral error can continue to integrate, which makes the controller sluggish to responding future changes in the opposite direction.
Describe the solution you would like.
I am not settled on a solution as yet. One possible solution is if the controller is saturated, set the reference output/state to their values at saturation or simply do not integrate. The former is known as conditioning (Hanus 1987) and the latter is "conditional integration" which is an old technique and ad hoc. The benefit, however, is that it would generalize to MIMO. See Astrom+Rudqwist 1989 for survey of such methods.
Describe alternatives you have considered.
Currently, the way antiwindup is handled in an ad hoc manner known as (gradual) back-calculation (Astrom+Rudqwist 1989, Astrom+Murray 2009). This "works" if set up appropriately, but is most applicable to SISO systems. In the library, there is a hack for expanding to MIMO, but this should really be avoided
Additional Context
N/A
The text was updated successfully, but these errors were encountered:
Is your feature request related to a problem? Please describe.
When actuator saturates, the integral error can continue to integrate, which makes the controller sluggish to responding future changes in the opposite direction.
Describe the solution you would like.
I am not settled on a solution as yet. One possible solution is if the controller is saturated, set the reference output/state to their values at saturation or simply do not integrate. The former is known as conditioning (Hanus 1987) and the latter is "conditional integration" which is an old technique and ad hoc. The benefit, however, is that it would generalize to MIMO. See Astrom+Rudqwist 1989 for survey of such methods.
Describe alternatives you have considered.
Currently, the way antiwindup is handled in an ad hoc manner known as (gradual) back-calculation (Astrom+Rudqwist 1989, Astrom+Murray 2009). This "works" if set up appropriately, but is most applicable to SISO systems. In the library, there is a hack for expanding to MIMO, but this should really be avoided
Additional Context
N/A
The text was updated successfully, but these errors were encountered: