1. The get_losses function itself is pretty clearly documented (see Section 
9.2.4 in the User’s Manual <http://www.pserc.cornell.edu/matpower/manual.pdf>) 
so hopefully there’s no confusion there. The LSF calculations in 
t/t_get_losses.m are specific to the power flow, not the OPF, in that they 
assume fixed generator voltages and a specific slack bus, etc.

2. There should be no difference, simply that get_losses() is more convenient 
than pulling code out of printpf() to compute it yourself.

3. In t/t_get_losses.m I compute the loss sensitivities numerically, by doing 
small (epsilon) perturbations of the load at each bus and recording the 
corresponding proportional change in losses. That’s what a “sensitivity factor” 
means. Then I compare this LSF matrix computed by numerical perturbation with 
the one I computed analytically to test that the latter is correct.

   Ray


> On Jan 20, 2015, at 12:22 PM, nivedita arunachalam 
> <[email protected]> wrote:
> 
> Dear Ray,
>              I have some doubts regarding the loss sensitivity factor 
> calculations you sent me. 
> 1. Is it for opf calculation alone?
> 2. I work with normal runpf operations and I get power loss by calling them 
> from printpf (sum(real losses)). Then what is the difference in your get loss 
> file and direct values.
> 3. Why do you use epsilon values in lsf caculation, what formula does it 
> represents?
> 
> Kindly reply me...
> 

Reply via email to