DcValueVoltageInitializer fails in presence of resistive only branches

[jeandemanged]

Signed-off-by: Damien Jeandemange damien.jeandemange@artelys.com

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 ?
No

What kind of change does this PR introduce?
Bug fix

What is the current behavior? (You can also link to an open issue here)
Exception thrown by DcValueVoltageInitializer if the newtwork contains a resistive only branch.

What is the new behavior (if this is a feature change)?
TBD

Does this PR introduce a breaking change or deprecate an API?
No

Other information:

java.lang.IllegalArgumentException: Branch 'l12' has reactance equal to zero

	at com.powsybl.openloadflow.dc.equations.AbstractClosedBranchDcFlowEquationTerm.<init>(AbstractClosedBranchDcFlowEquationTerm.java:42)
	at com.powsybl.openloadflow.dc.equations.ClosedBranchSide1DcFlowEquationTerm.<init>(ClosedBranchSide1DcFlowEquationTerm.java:24)
	at com.powsybl.openloadflow.dc.equations.ClosedBranchSide1DcFlowEquationTerm.create(ClosedBranchSide1DcFlowEquationTerm.java:33)
	at com.powsybl.openloadflow.dc.equations.DcEquationSystem.createImpedantBranch(DcEquationSystem.java:75)
	at com.powsybl.openloadflow.dc.equations.DcEquationSystem.createBranches(DcEquationSystem.java:116)
	at com.powsybl.openloadflow.dc.equations.DcEquationSystem.create(DcEquationSystem.java:126)
	at com.powsybl.openloadflow.dc.DcLoadFlowContext.getEquationSystem(DcLoadFlowContext.java:32)
	at com.powsybl.openloadflow.dc.DcLoadFlowEngine.run(DcLoadFlowEngine.java:65)
	at com.powsybl.openloadflow.dc.DcValueVoltageInitializer.prepare(DcValueVoltageInitializer.java:58)
	at com.powsybl.openloadflow.ac.DcValueVoltageInitializerTest.testFourBusNetworkResistiveOnlyBranch(DcValueVoltageInitializerTest.java:92)
``

[geofjamg]

In fact, in addition to the regression from #666, there is also several issues that were already before:

• spanning tree which is based in zero impedance criteria was only store for current dc mode, it should be available for both dc and ac mode in case of init DC + LF AC in order to have a consistent spanning tree with zero impedance booleans.
• non impedance branches graph and spanning tree should be cached as created several times (especially at the end for flow calculation on zero impedance branches)

So I tried in this PR to store all zero impedance infos for both AC and DC and as much as possible set everything in a lazy way to only calculate what is needed.

[geofjamg]

I think we should not do this here at a post loading fix. z cut could be done in a cleaner way at branch creation.

[annetill]

It has been designed like that because min z what dependent of the AC or DC context. Now, as we store both, we can do that at branch creation. Do you want me to do this?

[geofjamg]

Instead of the boolean dc which is not very readable we could have an enum like ?

enum EquationMode {
    AC,
    DC_APPROX
}

[geofjamg]

Maybe zeroImpedanceSubGraph and zeroImpedanceSpanningTree could be grouped in another object.

[annetill]

Yes, it could be a good idea.

[sonarqubecloud]

Kudos, SonarCloud Quality Gate passed!   

0 Bugs
0 Vulnerabilities
0 Security Hotspots
0 Code Smells

95.1% Coverage
0.0% Duplication