Mesbah-Lab-UCB / BO4Policy_Search_Plasma

Towards Personalized Plasma Medicine via Data-efficient Adaptation of Fast Deep Learning-based MPC Policies

BO4Policy_Search_Plasma

Contribution to the American Control Conference (ACC) 2023:

Towards Personalized Plasma Medicine via Data-efficient Adaptation of Fast Deep Learning-based MPC Policies

DOI

Authors: Kimberly J. Chan, Georgios Makrygiorgos, and Ali Mesbah

If you use our work, please cite:

@inproceedings{chan2023towards,
  title={Towards Personalized Plasma Medicine via Data-Efficient Adaptation of Fast Deep Learning-based {MPC} Policies},
  author={Chan, Kimberly J and Makrygiorgos, Georgios and Mesbah, Ali},
  booktitle={2023 American Control Conference (ACC)},
  pages={2769--2775},
  year={2023},
  organization={IEEE}
}

Implementation

To run this code on your own device, it is recommended to work within a virtual environment. You may create your own virtual environment and then install the required Python packages by using the command pip3 install -r requirements_sim_only.txt (for simulations only) and pip3 install -r appj_requirements.txt (additional packages for experiments with the cold atmospheric plasma jet (CAPJ).

The main file to run simulations with Bayesian optimization is src/run_sim.py.

The main file to run experiments with the in-house Mesbah Lab CAPJ testbed is src/run_exp.py

Additional details may be found within the src folder README as well as in commentary within the files.

Additional Dependencies for Sensitivity Analysis

This project performs a global sensitivity analysis (GSA) on the parameters of a deep neural network (DNN) to the closed-loop metrics of the plasma system (in simulation). For this portion of the project, we interface our Python scripts within MATLAB to take advantage of the established software package UQLab. The main script to run this analysis is src/sensitivity.m. To ensure that MATLAB is able to interface with Python, make sure to initialize your MATLAB session with pyenv('Version', 'PATH_TO_VENV'), where PATH_TO_VENV is the path to the python executable of the virtual environment created for this project. Furthermore, when installing a Python distribution for this purpose, the --enable-shared flag must be passed into the configuration for the installation. More details in the following links:

• MathWorks Documentation
• Stack Overflow Answer (links to below)
• Pyenv Documentation

(c) 2023 Mesbah Lab

in collaboration with George Makrygiorgos and Ali Mesbah.

Questions regarding this code may be directed to kchan45 (at) berkeley.edu