I would be interested in seeing the changes that you needed to make to
the power flow in order to incorporate FACTS devices, if you are
willing to share them. Both fmincon and the new pdipm solver handle
general problems of the form
min f(x)
x
subject to
h(x) = 0
g(x) <= 0
x_min <= x <= x_max
They both use the same functions for evaluating f(), g() and h() and
their derivatives. So if you begin your modifications with the current
fmincon solver, they should be directly applicable to the pdipm solver
in the next version of MATPOWER. Depending on what is involved, this
may actually be a good time to attempt to include modifications in
MATPOWER to either model FACTS devices or make it easier to do so.
Let's continue this discussion off-list.
--
Ray Zimmerman
Senior Research Associate
428-B Phillips Hall, Cornell University, Ithaca, NY 14853
phone: (607) 255-9645
On Apr 13, 2009, at 2:20 PM, bahrami wrote:
thanks a lot professor Zimmerman
yes I need to actually modify the problem formulation manually,I did
it for power flow algoritm in MATHPOWER and it was successful,now I
want to do it with OPF
you said " we have a pure-Matlab version of the PDIPM solver in the
works" how can I get it?
I know what I need for modelling of FACTS devices,but this is very
time consuming and I want to get best result,so I decided to spend
my time for modifying the best algoritm(PDIPM)
thanks for help me.but if it's possible to help me again,please tell
me what do I do?which algoritm should I choose to start?
thanks,thanks,thanks....
--- On Mon, 4/13/09, Ray Zimmerman <[email protected]> wrote:
From: Ray Zimmerman <[email protected]>
Subject: Re: modify_opf
To: "MATPOWER discussion forum" <[email protected]>
Date: Monday, April 13, 2009, 10:22 PM
In order to modify the pdipmopf MEX file you would need the C
language sources, which we have not make publicly available. The
first question you need to answer is, how are the FACTS devices to
be modeled? And how does that affect the overall problem
formulation? Does it involve additional linear constraints?
Modification of the existing non-linear constraints?
If it's the former you can simply use the generalized formulation to
add the constraints as described in the manual and the existing
algorithms should be able to solve it without modification. If it is
the latter, you will need to actually modify the problem formulation
manually. In this case, I would suggest modifying the fmincon-based
solver, since it is pure-Matlab. The fmincon algorithm is not
suitable for large systems, but we have a pure-Matlab version of the
PDIPM solver in the works that uses the same cost and constraint
evaluation functions and it is able to handle large systems. The
only difference is that it requires 2nd derivative information for
all of the costs and constraints.
Hope this helps,
--
Ray Zimmerman
Senior Research Associate
428-B Phillips Hall, Cornell University, Ithaca, NY 14853
phone: (607) 255-9645
On Apr 12, 2009, at 3:27 AM, bahrami wrote:
Dear professor Zimmerman
i want to modify opf for modeling of FACTS devices
i want to begin with "pdipmopf" algoritm
1- is it posible to modify this algoritm,thogh it's a mex file
2- is there any alternative algoritm to select for modify?
3-i think "pdipmopf" algoritm is the best algoritm for this,is'nt it?
pleas help me to start this project
thanks
