An upper-level agent makes generation and transmission line investment decisions, while the market is cleared in the lower-level. The lower-level problem is a robust optimization problem in which some parameters are stochastic.
Tested with Python 3* and Gurobi 9*.
In clustering.py, select N_CLUSTERS (operating conditions). Generate operating conditions by running
python clustering.py
In common_data.py, set num_scenarios to match the selected number of operating conditions. Set other flags to modify the model behavior. Call robust.py with the selected master problem and subproblem algorithm:
python robust.py <benders_dc|milp_dc> <miqp_dc|milp_dc> <output_dir>
Using the output files of robust.py as input, generate plots and tables in the manuscript by running
python cost_of_robustness2.py <list of costs, e.g. 100 1000 10000>
and plotting.py, generation_mix*.py, summary.py, etc.