Roadmap

[annetill]

Describe the current behavior

Java

• Are we resilient to a lack of high and low voltage limit on a voltage level? No, to be fixed in Java code.
• Connectivity computation could be transferred in Java.
• Maximum slack reactive total amount by bus of '500MVar' should be configurable.
• Add unit test with a voltage level with a missing high or low voltage limit.
• Generator with P = 0:
• Reporter.
• List of generators with constant targetQ and targetV.
• Idea: AMPL starter.

# Units up and generating:
# Warning: units with Ptarget=0 are considered as out of order

AMPL

• We have constraint in AMPL part that check low voltage limit, high voltage limit and the "ideal" ratio. This constraint is a check done before any computation. This check could be transferred in the Java part.
  1) Do we need to keep this constraint.
  2) Do we need to fix this constraint?
• MIN_GENERATION: ...
• Clarification of PQ vs PV buses. The status voltageRegulationOn is not used by the optimizer part. But we give to the simualtor a list of generators with constant targetQ.
• Clarify management of generator with P = 0.
• Log management.
• Split the .run and .mod files in several ones.
• Transfer the documentation in the README.md.
• AMPL Outputs written in JSON.

Describe the expected behavior

To be discuss

Describe the motivation

No response

Extra Information

No response

[p-arvy]

In addition to the points mentioned above (most of which have been completed), I note the following points to be added to the ROADMAP.

AMPL

• Split the code to isolate functional blocks and test them as a unit.
• Transfer some of the blocks in Java (in particular Connex Component calculation and DCOPF).
• Hexagonal reactive power bounds for generators.

Java

• Correct the per unitage of lines connecting two different nominal voltages in AMPL export.
• Make test resources independent of execution time/place (see Make test resources independent of execution time/place #63 )

Exploratory

• Addition of discrete control for shunts/transformers in ACOPF. Same for line opening.
• Optimization of ACOPF via a workflow splitted into several optimizations.

Do not hesitate to complete @Hadrien-Godard @klorel @So-Fras