Scenario Based robust control

Run and tested in Python 3.9.12

Edit options.py to choose PRISM max memory usage, and enter PRISM path

use run.py to execute

For full details, see paper.pdf

Dependencies

• Python packages:
  • numpy 1.23.4
  • tqdm 4.64.1
  • matplotlib 3.5.3
  • scipy 1.9.3
• Java Development Kit
• PRISM (iMDP branch) - https://github.com/davexparker/prism

To reproduce experiments from the paper:

• UAV Plot: select UAV_var_noise with medium noise
• Room heating (main): select steered_n_room_heating with 2 rooms and medium noise
• Room heating (unsteered): select n_room_heating with 2 rooms and medium noise
• More general formula: Run test.py

To Do:

• Bash file to run all experiments
• Make parameters more user friendly
• Tidy code
• Ensure all options work