Prev: How I can generate the training data set for multiple inputs in ANFIS toolbox
Next: Problems with fmincon when trying to compute transition probabilities
From: Kerose Woldetsadik on 30 Jun 2010 12:07
I was designing " Chattering free sliding mode speed control for Induction motor drive" . While building the simulink model for the controller i came across one problem , that is constructing eqn(1) using simulink block( not S-function). If any one help me with that my project is over.

(1) S(t)= e(t) + ∫e(ζ)dζ-e(0)................Sliding surface for Total

In other word how can I implement e(0) using just simulink blocks
where e(t)= wr(t)-wr*(t)
wr(t)= feedback speed
wr*(t)= reference speed
Thank you for giving time to read my request
Sincerely
DMS(ethiopia)
From: Frank W. on 1 Jul 2010 06:23
> In other word how can I implement e(0) using just simulink blocks
Set either the 'Inital condition' of the integrator block to the desired start value or use the 'IC' (initial condition) block from Simulinks Signal Attributes library section.
From: Kerose Woldetsadik on 1 Jul 2010 06:40
"Frank W." <yohansan.nospam(a)web.de> wrote in message <i0hqa8$pqu$1(a)fred.mathworks.com>...
> > In other word how can I implement e(0) using just simulink blocks
> Set either the 'Inital condition' of the integrator block to the desired start value or use the 'IC' (initial condition) block from Simulinks Signal Attributes library section.

hey Frank W.,
It's really nice to have a prompt response. The thing is I can't just set e(0) to any desired value, it should be the exact value of e(t) at t=0. which is different from initial condition(IC). Is there any way to go around this problem using only Simulink blocks?
Any help is appreciated.

where: e(t)=wr(t)-wr*(t)
wr(t)= actual feedback speed signal
wr*(t)= reference speed signal
sincerely
DMS(from Ethiopia)
From: Frank W. on 2 Jul 2010 04:40
> It's really nice to have a prompt response. The thing is I can't just set e(0) to any >desired value, it should be the exact value of e(t) at t=0. which is different from initial >condition(IC).
The intial condition value is the one at simulation start t=0. I assume you know the value e(t=0) because you set the simulation up. Or is your problem that you do not know the initial value?
 | 
Pages: 1
Prev: How I can generate the training data set for multiple inputs in ANFIS toolbox
Next: Problems with fmincon when trying to compute transition probabilities