The time-domain control techniques in general, and Sliding Mode Control (SMC) in particular, do not care about transient response specifications (Percent Overshoot, Rise Time, Settling Time, etc.) that frequency-domain techniques account well for. Most of the references in SMC, use "trial-and-error" to get the desired specifications, and focus on steady-state specs. I would like to know how SMCs account for these specifications in order to be able to compare them to other control techniques.
Hi Michel
The most common parameter in the conventional SMC is reaching time (when sliding variable reaches to zero), and there are not much discussions about overshoot or rise time.
But by defining different sliding surfaces (PI or PID) you will be able to have overshoot or ... . However mostly this type of definitions are not needed (under matched condition), because the conventional SMC have better performance.
Best Regards
The overshooting in reaching phase of a SMC can be controlled in the stage of defining the sliding surface " s=(d/dt + h)n-1*e " specifically by choosing proper h which will have its effect also on stability time and the chattering phenomena that accompanies the regular SMC.
I will be glad if you make a look at my publication for more clarification
Article Off-line tuning for adaptive sliding control using particle ...
Thank you Sajjad Shoja Majidabad, I do need these specifications because I am comparing the robustness of these techniques with frequency domain ones.
Dear Viswanath Devan,
Thank you for your answer, it is quite helpful, but I still need to relate the parameters' choice in SMC to the specifications.
Thank you Marwan Abdel Razik for your answer and paper. I will go over it and hopefully I can relate the h parameter with the specs.
anytime Michel Owayjan...good luck in your researches..
Actually it is lambda not h :)
My question is regarding the answer given by Viswanath Devan.....is there any command in MATLAB to find out rise time, overshoot, settling time etc. of SMC or we have to calculate those parameters from the obtained plot directly?
Dear Michel Owayjan,
The choice of sliding variable ensures the desired closed loop dynamics
Dear Ali,
Thank you for your feedback. I was looking for a relation between the closed loop dynamics (transient specs) and the choice of the variables in SMC (lambda...)
Dear Michel Owayjan,
You may get assistance from chapter 6 of the attached book
- Badges
- Science topic