Thanks once again for the response and clarifications.
Keep up your quality work!
---
Themistoklis C. Xygkis,
Dr Electr. & Comp. Eng. NTUA
Automation Systems MSc
Στις 2022-01-13 18:50, Ray Daniel Zimmerman έγραψε:
As I mentioned, the only change that should affect the results at all
is the updating of the generator outputs from the estimated solution.
Here is the discussion [1] that prompted that change. This should have
absolutely _no_ effect on the convergence or on the estimated state
found as the solution. Other than that, the changes were only to keep
the code compatible with updated function interfaces. I.e. no
functional changes.
So functionally, in terms of the algorithm and the voltage solutions
found, there shouldn’t _any_ changes from 4.1 to the current
version.
But you do have to use compatible versions of MATPOWER and your SE
code. Specifically, the call to dSbus_dV() at line 102 of doSE.m [2]
must return [dSbus_dVm, dSbus_dVa] if used with MATPOWER 6 and
earlier, and [dSbus_dVa, dSbus_dVm] if used with MATPOWER 7 and later.
That should be the only incompatibility between versions.
Ray
On Jan 8, 2022, at 10:35 AM, Themistoklis Xygkis
<txyg...@power.ece.ntua.gr> wrote:
Thank you for your response.
I have used MATPOWER4.1 for a wide range of test systems (benchmark,
real-world etc.) and the related results have been highly consistent
and reliable, compared to the ones ontained from other academic SE
tools.
So, what was the need for the changes made in the current MATPOWER
version compared to the older ones?
In what way do they improve the MATPOWER?
Thank you once again.
---
Themistoklis C. Xygkis,
Dr Electr. & Comp. Eng. NTUA
Automation Systems MSc
Στις 2022-01-08 00:25, Ray Daniel Zimmerman έγραψε:
You will definitely encounter problems if trying to use SE code from
before MATPOWER 7 with MATPOWER 7 or later … or vice-versa, since
there was an incompatible change introduced in v7 to the dSbus_dV()
function called in doSE() (order of output args was swapped).
In general, you want to make sure that the SE code you are using is
what is found in the extras/se sub-directory of the MATPOWER version
you are running it under.
Other than that change, the only changes I see that should have any
effect on the result is the section on updating Pg and Qg in the
solution using estimated values at [1]lines 55 [2]-64 in run_se.m.
Hope this helps,
Ray
On Jan 7, 2022, at 10:24 AM, Themistoklis Xygkis
<txyg...@power.ece.ntua.gr> wrote:
Dear colleagues,
I experience some "compatibility" issues while I am trying to run
the state estimation (SE) code I had developped during my PhD
research, in newer Matlab versions. My SE code was based on the doSE
function of MATPOWER 4.1 and all my simulations had been conducted
in Matlab 2012b.
Currently, I am trying to run this code in Matlab 2018b; I use some
standard IEEE systems which I transform into mpc format and, then,
introduce into the SE code. Although the SE runs problem-free for
all systems in the original MATPOWER4.1 - Matlab 2012b environment,
I experience computational issues due to close-to-singular matrices
while I am trying to run the code in MATPOWER4.1 - Matlab 2018b
environment. Yet, this may be anticipated
due to singularity conditions which changed in newer Matlab
versions.
The problems get serious when I am trying to run the code in
MATPOWER7.0 - Matlab 2018b environment; convergence issues occur
and, of course,
I get results which are definitely wrong. So, given that my SE code
is based on the MATPOWER4.1 - doSE function, is it incompatible with
the functions of MATPOWER7.0; As I saw, the MATPOWER7.0- doSE
function has been changed due to the introduction of the section
"compute net injection at generator buses". Does this modification
or another that I cannont see, makes my SE code incompatible with
the latest MATPOWER?
Thank you in advance.
--
Themistoklis C. Xygkis,
Dr Electr. & Comp. Eng. NTUA
Automation Systems MSc
Links:
------
[1]
https://github.com/MATPOWER/mx-se/blob/cd7032e95930ae79a631b6fafed27351ccc35b0c/run_se.m#L55-L64
[2]
https://github.com/MATPOWER/mx-se/blob/cd7032e95930ae79a631b6fafed27351ccc35b0c/run_se.m#L55
Links:
------
[1] https://www.mail-archive.com/matpower-l@cornell.edu/msg05513.html
[2]
https://github.com/MATPOWER/mx-se/blob/cd7032e95930ae79a631b6fafed27351ccc35b0c/doSE.m#L102
