Employing the terminal slidning mode technique (S. Yu, X. Yu, et.al., doi: Article Continuous Infinite time control for robotic manipulators wi...
) for the second-order uncertain system to achieve finite-time or fixed-time stability, the control input and the velocity state will overshoot by a large amount during the system evolution. See the attached graph for details.Is there a effective scheme to avoid such phenomenon?
Siamak Heidarzadeh
: Hi, Siamak Heidarzadeh. Thanks for your answer. What I mean is that such problem arises when construct the control algorithm based on the terminal sliding mode surface presented in this paper, instead of directly implementing the controller in this paper.By the way, in my designed controller, the external disurbances is suppressed by the discontiguous term "-Ksign(s)", where "K" is the pisitive gain, "s" is the sliding mode vector, and this is different in the this paper. Then, could this be the cause for the problem?
Hello Chang-Duo Liang
It is also a unavoidable problem that took place in my desgining controller. From my perpective point, based on an observer, the positive gain "K" can be devised as a time-varying gain. More specifically, when the observer has been completely conducted, the positive gain "K" can be assigned a sufficiently small value to reduce the control overshoot.
However, it is still a difficult problem for me to relieve the control overshoot before the observer takes effect. I sincerely hope that this answer can help you effecively.
- Badges
- Science topic
- Similar topics
- Engineering
- Control Theory
- Control Systems
- Control Systems Engineering