Refactor shared voltage control modelling #408

geofjamg · 2021-12-26T20:36:15Z

Signed-off-by: Geoffroy Jamgotchian [email protected]

Please check if the PR fulfills these requirements (please use '[x]' to check the checkboxes, or submit the PR and then click the checkboxes)

The commit message follows our guidelines
Tests for the changes have been added (for bug fixes / features)
Docs have been added / updated (for bug fixes / features)

Does this PR already have an issue describing the problem ? If so, link to this issue using '#XXX' and skip the rest
No

What kind of change does this PR introduce? (Bug fix, feature, docs update, ...)
Refactoring

What is the current behavior? (You can also link to an open issue here)
Shared voltage control is modeled using one PQV bus at controlled bus and (Ngen - 1) reactive distribution equations at all controller buses except one. These equations allows to define reactive power distribution across all voltage controller buses.
The problem with such a modeling is that when one controller bus of the shared control switch PQ (reactive limit bound reached), as reactive power equations have been defined 2 by 2 using an arbitrary reference controller bus we have to remove all reactive power distribution equations and re-create all of them using potentially a new reference controller bus.
This is a heavy operation and it should be simpler if we can just active/deactivate some predefined equations. Furthermore for later performance optimization we will try to define alternative equations (i.e couple of equation that when one is activate the other is deactivate automatically and vice versa) so that we only have to update Jacobian matrix when a bus is switched PV/PQ ou PQ/PV and not fully rebuild it.

TODO: same for transformer voltage control has to be done in this PR

What is the new behavior (if this is a feature change)?
We only need to deactivate/activate right predefined equations when switch a PV to PQ or PQ to PV bus. There is no more equation removal.

Does this PR introduce a breaking change or deprecate an API? If yes, check the following:

The Breaking Change or Deprecated label has been added
The migration guide has been updated in the github wiki (What changes might users need to make in their application due to this PR?)

Other information:

(if any of the questions/checkboxes don't apply, please delete them entirely)

Signed-off-by: Geoffroy Jamgotchian <[email protected]>

annetill · 2021-12-28T15:23:42Z

src/main/java/com/powsybl/openloadflow/ac/equations/AcEquationSystem.java

@@ -72,8 +65,7 @@ private static void createGeneratorControlEquations(LfBus bus, LfNetworkParamete
            if (voltageControl.isVoltageControlLocal()) {
                createLocalVoltageControlEquation(bus, networkParameters, equationSystem, creationParameters);
            } else if (bus.isVoltageControlled()) {
-                // remote controlled: set voltage equation on this controlled bus
-                createVoltageControlledBusEquations(voltageControl, networkParameters, equationSystem, creationParameters);
+                createRemoteVoltageControlEquations(voltageControl, networkParameters, equationSystem, creationParameters);


If I remember well we have also a case where we create distribution equations for remote and local controllers. We fall in that second case. You can keep remote in the name of the method, but maybe I can add it in the comment.

Looking at the code there is nothing specific to local controller even if it works with hybride remote/local controller. The only local mode is handled in the previous if.

annetill · 2021-12-28T15:28:37Z

src/main/java/com/powsybl/openloadflow/network/LfBus.java

@@ -152,4 +152,8 @@ default boolean isParticipating() {
     * Find bus + parallel branches neighbors.
     */
    Map<LfBus, List<LfBranch>> findNeighbors();
+
+    double getRemoteControlReactivePercent();


Can you add voltage in the name of the method because we also have shunt remote voltage control and reactive remote control.

annetill · 2021-12-28T15:41:41Z

src/main/java/com/powsybl/openloadflow/network/VoltageControl.java

+
+        // update bus reactive keys
+        double reactiveKeysSum = Arrays.stream(reactiveKeys).sum();
+        for (int i = 0; i < controllerBuses.size(); i++) {


How do you know here that the sum is never equal to zero after updating the keys?

Yes it might happen, when all generators of the control, go PQ.
=> I added a check

annetill

Maybe just some changes/comments, but the idea is good and looks right. I have run the LF on all my real test cases (IOP, RGCE, matpower cases) and all is okay. But during my test on a TYNDP file, I fall in an exception Caused by: com.powsybl.commons.PowsyblException: Generators [_9064117a-2da4-46db-b1ee-f60810974a76, _ee520ab7-786a-499b-a3ae-d96fb5b12596] connected to bus '_55207717-1417-458c-9b5f-e67a3eedc46a_0' must control the voltage of the same bus. Can we do something for that and just log a warn or do we assume the exception? I don't remember this issue with this file indeed.

geofjamg · 2021-12-28T17:00:07Z

Do you also have it on main branch?

annetill · 2021-12-28T17:12:54Z

Do you also have it on main branch?

Yes. I have planned to work on stand by automaton of SVC tonight and tomorrow. And I think it will be better to have this PR merged before... even if it means a new rebase for the PR of shunt voltage control ! Do you think it is possible?

Signed-off-by: Geoffroy Jamgotchian <[email protected]>

geofjamg · 2021-12-28T23:45:50Z

Maybe just some changes/comments, but the idea is good and looks right. I have run the LF on all my real test cases (IOP, RGCE, matpower cases) and all is okay. But during my test on a TYNDP file, I fall in an exception Caused by: com.powsybl.commons.PowsyblException: Generators [_9064117a-2da4-46db-b1ee-f60810974a76, _ee520ab7-786a-499b-a3ae-d96fb5b12596] connected to bus '_55207717-1417-458c-9b5f-e67a3eedc46a_0' must control the voltage of the same bus. Can we do something for that and just log a warn or do we assume the exception? I don't remember this issue with this file indeed.

This is manageable but we have probably to rethink our controller modeling. A controller bus is supposed in our design, to be part of only one voltage control (only one voltageController attribute in AbstractLfBus for instance). So there is quite a lot of rework to support it. Let's create an issue and do it in another PR?

What I do not understand is how we didn't had this case before. Is it something new in our test cases? Or do we just not skip it before?

geofjamg · 2021-12-28T23:47:54Z

Do you also have it on main branch?

Yes. I have planned to work on stand by automaton of SVC tonight and tomorrow. And I think it will be better to have this PR merged before... even if it means a new rebase for the PR of shunt voltage control ! Do you think it is possible?

I'm not 100% confident with this PR. Is it really linked to SVC devs? Maybe you could just start from this branch to easily rebase when merged.

Signed-off-by: Geoffroy Jamgotchian <[email protected]>

sonarqubecloud · 2021-12-29T19:52:57Z

Kudos, SonarCloud Quality Gate passed!

0 Bugs
0 Vulnerabilities
0 Security Hotspots
0 Code Smells

94.7% Coverage
0.0% Duplication

geofjamg · 2021-12-29T21:37:01Z

Do you also have it on main branch?

Yes. I have planned to work on stand by automaton of SVC tonight and tomorrow. And I think it will be better to have this PR merged before... even if it means a new rebase for the PR of shunt voltage control ! Do you think it is possible?

I'm not 100% confident with this PR. Is it really linked to SVC devs? Maybe you could just start from this branch to easily rebase when merged.

Looks good in fact.

annetill

It looks good indeed. Do you want me to run all the tests again?

geofjamg · 2021-12-30T08:41:23Z

It looks good indeed. Do you want me to run all the tests again?

If you already did it, no.

Refactor shared control modelling

d02b736

Signed-off-by: Geoffroy Jamgotchian <[email protected]>

geofjamg changed the title ~~Refactor shared control modelling~~ [WIP] Refactor shared control modelling Dec 26, 2021

geofjamg added 2 commits December 27, 2021 15:30

Fix

f004270

Signed-off-by: Geoffroy Jamgotchian <[email protected]>

Fix

89221f2

Signed-off-by: Geoffroy Jamgotchian <[email protected]>

geofjamg changed the title ~~[WIP] Refactor shared control modelling~~ [WIP] Refactor shared voltage control modelling Dec 27, 2021

geofjamg requested a review from annetill December 27, 2021 21:51

geofjamg added 4 commits December 27, 2021 22:58

Cleanup

ccca33a

Signed-off-by: Geoffroy Jamgotchian <[email protected]>

Add doc

9a2bd7c

Signed-off-by: Geoffroy Jamgotchian <[email protected]>

Fix Sonar issue

6dfe895

Signed-off-by: Geoffroy Jamgotchian <[email protected]>

Fix

ee4e8db

Signed-off-by: Geoffroy Jamgotchian <[email protected]>

annetill reviewed Dec 28, 2021

View reviewed changes

annetill requested changes Dec 28, 2021

View reviewed changes

Review fixes

80da686

Signed-off-by: Geoffroy Jamgotchian <[email protected]>

geofjamg added 2 commits December 29, 2021 20:49

Clean slope

74b3272

Signed-off-by: Geoffroy Jamgotchian <[email protected]>

Clean

01aa8ce

Signed-off-by: Geoffroy Jamgotchian <[email protected]>

geofjamg changed the title ~~[WIP] Refactor shared voltage control modelling~~ Refactor shared voltage control modelling Dec 29, 2021

geofjamg requested a review from annetill December 29, 2021 21:38

annetill approved these changes Dec 30, 2021

View reviewed changes

annetill merged commit 5dbd806 into main Dec 30, 2021

annetill deleted the refactor_shared_control_2 branch December 30, 2021 08:46

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Refactor shared voltage control modelling #408

Refactor shared voltage control modelling #408

geofjamg commented Dec 26, 2021 •

edited

Loading

annetill Dec 28, 2021

geofjamg Dec 28, 2021

annetill Dec 28, 2021

geofjamg Dec 28, 2021

annetill Dec 28, 2021

geofjamg Dec 28, 2021

annetill left a comment

geofjamg commented Dec 28, 2021

annetill commented Dec 28, 2021

geofjamg commented Dec 28, 2021

geofjamg commented Dec 28, 2021

sonarqubecloud bot commented Dec 29, 2021

geofjamg commented Dec 29, 2021

annetill left a comment

geofjamg commented Dec 30, 2021

Refactor shared voltage control modelling #408

Refactor shared voltage control modelling #408

Conversation

geofjamg commented Dec 26, 2021 • edited Loading

annetill Dec 28, 2021

Choose a reason for hiding this comment

geofjamg Dec 28, 2021

Choose a reason for hiding this comment

annetill Dec 28, 2021

Choose a reason for hiding this comment

geofjamg Dec 28, 2021

Choose a reason for hiding this comment

annetill Dec 28, 2021

Choose a reason for hiding this comment

geofjamg Dec 28, 2021

Choose a reason for hiding this comment

annetill left a comment

Choose a reason for hiding this comment

geofjamg commented Dec 28, 2021

annetill commented Dec 28, 2021

geofjamg commented Dec 28, 2021

geofjamg commented Dec 28, 2021

sonarqubecloud bot commented Dec 29, 2021

geofjamg commented Dec 29, 2021

annetill left a comment

Choose a reason for hiding this comment

geofjamg commented Dec 30, 2021

geofjamg commented Dec 26, 2021 •

edited

Loading