Skip to content

About

Agentic AI for power-system operators: presentations, hands-on Colab labs and source-backed examples.

Resources

Stars

5 stars

Watchers

0 watching

Forks

Latest commit

 

History

27 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Agentic AI for power-system operators

Learn the concepts and follow three operator cases before trying the optional notebooks.

Alaa Selim, MIT/Dartmouth · Mohamed Chehada, UT Austin

Open the tutorial website

1. View the presentations

Use the arrow keys to move through the slides. Source credits accompany each presentation.

2. Optional practice after the case tutorial

One notebook covers all three hands-on cases.

Open Classroom Labs in Colab

  1. Save a copy in Google Drive and connect a CPU runtime.
  2. Run Set up Google Colab, wait for Ready, then start at L0.
  3. Continue through L12, including L10B. Keep RUN_OPENDSS and RUN_LOCAL_LLAMA set to False.

No GPU, API key or language-model download is needed for the standard lab. Setup requires internet access.

3. Work through the cases

  • Voltage · L2-L5: compare a plan and a saved learned controller. Change the current allowance and recheck the result.
  • Restoration · L6-L7: compare total and critical load served as tie capacity changes.
  • Cyber · L8-L9: compare an isolated error and a coordinated attack, then inspect trusted replacement measurements.
  • Agent choices · L10-L10B: inspect saved Llama replies, tool results and next steps.

Run L0-L1 first. Change one exercise input at a time and reset it before continuing. Setup and troubleshooting.

Downloads and references

Explore further

The concept tutorial stands on its own. Classroom Labs and these companions are separate practice activities:

Read results in context

The cases use synthetic or reduced networks. Python computes fresh numerical results; saved model replies are labeled separately. L10B makes no new model call or electrical solve. Results apply to the stated model and assumptions. No lab operates equipment.

Source projects and earlier editions

Earlier editions remain in their original folders for reproducibility. Start with the current presentations and Classroom Labs above.

About

Agentic AI for power-system operators: presentations, hands-on Colab labs and source-backed examples.

Resources

Stars

5 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages