This project develops supervised neural surrogate models to learn the DC-OPF economic dispatch operator for 24-hour flexible data center scheduling. Trained on CVXPY-GUROBI solutions of the IEEE-14 system, the models achieve 450–700× faster inference with <0.2% cost deviation, enabling real-time grid-constrained dispatch emulation.
-
Notifications
You must be signed in to change notification settings - Fork 0
This project develops supervised neural surrogate models to learn the DC-OPF economic dispatch operator for 24-hour flexible data center scheduling. Trained on CVXPY-GUROBI solutions of the IEEE-14 system, the models achieve 450–700× faster inference with <0.2% cost deviation, enabling real-time grid-constrained dispatch emulation.
ashutoshshirole/NeuralED
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
About
This project develops supervised neural surrogate models to learn the DC-OPF economic dispatch operator for 24-hour flexible data center scheduling. Trained on CVXPY-GUROBI solutions of the IEEE-14 system, the models achieve 450–700× faster inference with <0.2% cost deviation, enabling real-time grid-constrained dispatch emulation.
Resources
Stars
Watchers
Forks
Releases
No releases published