Hi Xian,

It is true that makePTDF() 
<http://www.pserc.cornell.edu/matpower/docs/ref/matpower6.0/makePTDF.html> is 
not suitable for large systems, since it does compute sensitivities of flows in 
all branches with respect to injections at all buses. For a large system, if 
you want to compute the sensitivities of flows to a particular transfer (e.g. 
between bus k and the slack), you can do it by solving …

B_dc * ∆x = ∆P_dc

… for ∆x, based on eq (4.7) in the User’s Manual, where ∆P_dc is constructed 
according to the desired transfer. Then compute the corresponding ∆P_f from 
(3.29). You can use makeBdc() 
<http://www.pserc.cornell.edu/matpower/docs/ref/matpower6.0/makeBdc.html> to 
construct B_dc and B_f.

If you needed to do this for multiple transfers, you could factor B_dc once 
beforehand and use the factors to efficiently solve for each ∆x.

The option to compute these sensitivities this way would be a nice addition to 
makePTDF() 
<http://www.pserc.cornell.edu/matpower/docs/ref/matpower6.0/makePTDF.html> and 
I would be happy to include such a patch if anyone has the time to create it 
(see the MATPOWER Contributor’s Guide 
<https://github.com/MATPOWER/matpower/blob/master/CONTRIBUTING.md>). Feel free 
to add this to the issue tracker <https://github.com/MATPOWER/matpower/issues> 
as an enhancement request (and reference this e-mail) if you don’t have to do 
it yourself, so maybe someone else will take it on.

Hope this helps,

   Ray




> On Jan 31, 2017, at 1:46 AM, 郭贤 <[email protected]> wrote:
> 
> Dear all,
> 
> The "make PTDF" is to calculate sensitivity factors for all the buses, since 
> I have big system with around 20000 buses, when I try to use "make PTDF" 
> function, the MATLAB program got stuck. Since I did not need PTDF for all 
> buses, I'm wondering how could I have PTDF for partial buses without 
> calculating PTDF for all the buses?
> 
> 
> Thanks a lot. I really appreciate it. 
> 
> -- 
> —————————
> Thanks & Regards
> Xian Guo
> Iowa State University 
> Research Assistant in Electric Power and Energy System
> ECpE
> 1207 Coover Hall

Reply via email to