Skip to content

Antares optimization problem retrieval

In this step, antares-solver runs the Antares study with a specific configuration set by Antares-Xpansion and writes the weekly optimization problems as .mps files. More precisely, antares-solver performs two successive optimizations, see Antares-Simulator reference guide:

  1. A linear relaxation of the Antares MILP problem is solved where the on/off variables of thermal clusters are continuous.
  2. A heuristic search for integer values of these variables, satisfying the dynamic constraint of the system with the smallest possible cost is then carried out. Finally, a second linear optimization with the integer variables set to the previously found values is performed.

In Antares-Xpansion, the weekly problems solved in the first optimization step are written as .mps files, the files generated by the second optimization are discarded. Some additional files (variable.txt, area.txt and interco.txt) are also generated to map the variables of the Antares optimization problems to their phhysical meaning.

Specific configuration for Antares-Xpansion

In order to run the Antares study correctly, some specific simulation options are required. The Python orchestrator changes the file settings/generaldata.ini, with some settings depending on the uc_type parameter defined in the user/expansion/settings.ini file.

  • If uc_type = expansion_fast, the following settings are used:

    [optimization]
    include-exportmps = true
    include-exportstructure = true
    include-tc-minstablepower = false
    include-tc-min-ud-time = false
    include-dayahead = false
    [general]
    mode = Economy
    [other preferences]
    unit-commitment-mode = fast
    
  • If uc_type = expansion_accurate, the following settings are used:

    [optimization]
    include-exportmps = true
    include-exportstructure = true
    include-tc-minstablepower = true
    include-tc-min-ud-time = true
    include-dayahead = true
    [general]
    mode = expansion
    [other preferences]
    unit-commitment-mode = accurate
    
Parameters Remark
include-exportmps Ask for export of .mps files.
include-exportstructure Ask for export of structure files : variables.txt, area.txt, interco.txt.
include-tc-minstablepower Enable constraints on minimum stable power for thermal clusters.
include-tc-min-ud-timer Enable constraints on minimum up and down times for thermal clusters.
include-dayahead Enable day ahead reserves (i.e. limitation of the maximum production power in order to be able to react to infraday unexpected events).
mode Sets the Antares simulation mode.
unit-commitment-mode Sets the Antares unit-commitment mode, see the unit-commitment mode of Antares-Simulator.

Output files

After the Antares simulation, some output files of interest for Antares-Xpansion are generated.

.mps files for the weekly optimization problems

For each week and Monte-Carlo year, one .mps file is generated, encoding the linear problem that is solved by Antares-Simulator.

File name formatting: problem-<mc_year>-<week>-<timestamp>.mps.

Example

problem-1-1-20210928-155703.mps represents the weekly problem for MC year 1 and week 1 generated on 2021/09/28 at 15h57m03s.

variables.txt file for each .mps file

Each .mps file has a corresponding variable.txt file that allows to map each variable number of the .mps file to its physical representation.

File name formatting :variables-<mc_year>-<week>-<timestamp>.txt

Example

variables-1-1-20210928-155703.txt represents the variable names of the problem for MC year 1 and week 1 generated on 2021/09/28 at 15h57m 03s.

The variable.txt file contains one line for each variable (i.e. column) in the weekly problem encoded in the corresponding .mps file. In the .mps file the variables are named Cxxxxxxx where xxxxxxx is the number of the variable.

The line content depends on the type of variable (if it is related to a country or not).

Example

For 462 ValeurDeNTCOrigineVersExtremite 0 0 14:

Value Parameter
462 Column ID in the .mps weekly problem.
ValeurDeNTCOrigineVersExtremite Column physical name.
0 Country.
0 Link ID.
14 Time step.

As the weekly problem is solved with an hourly time step, each physical variable is present 168 times. The time steps are numbered with reference to the first hour of the year.

Example

For the problem of week 2, the variable ValeurDeNTCOrigineVersExtremite 0 0 is present 168 times, with time steps ranging from 168 to 335.

The file area.txt

One file that contains the name of the areas of the Antares study is produced by antares-solver.

File name formatting: area-<mc_year>-<week>-<timestamp>.txt.

Because the areas do not change in each MC year or week, the file is created only for the first MC year and week simulated. The file contains one line for each area.

Example

area1
area2
flex
peak
pv
semibase

The file interco.txt

One file that describes the links of the Antares study is produced by antares-solver.

File name formatting: interco-<mc_year>-<week>-<timestamp>.txt.

Because the links do not change in each MC year or week, the file is created only for the first MC year and week simulated. The file contains one line for each link.

Example

The line

3 1 2
in interco.txt means that
link_id = 3 
index_area_origin = 1
index_area_extremite = 2

The index is a match to the index in the area.txt file. So for this example, the link_id 3 matches the link between area2 (area id 1) and flex (area id 2).