Skip to main content

Assured AI for the Decarbonization Era: Structural Designs on Deep Neural Network

Sponsored by

This event is over.

Event Details:

Distributed energy resources (DERs) have changed the landscape for real-time monitoring of distribution networks. However, the low observability in the secondary distribution grids makes monitoring hard, due to limited investment in the past and the vast coverage of distribution grids. To recover measurements for robustness, past methods proposed various machine learning models with some explainability. However, critical energy infrastructure needs assurance. In this talk, we will discuss how to build and measure assurance via a structured deep neural network. The topics cover maximization of physical interpretability, inverse neural networks for state estimation, and heterogeneous transfer learning for new infrastructures. To verify these methods for assurance, we will show how to build software for extensive validation, establish hardware-in-the-loop validation, and create GIS-based validation with scalability. The results show assurance for using the proposed machine learning methods, providing a cost-effective way for deep decarbonization.

The seminar is open to the public. Stanford students enrolled in CEE 272T/EE 292T: SmartGrids and Advanced Power Systems Seminar must attend in person. All others, please register to attend on Zoom via the RSVP link.

Stream Information:

Join Stream

Dial-In Information

Join from a PC, Mac, iPad, iPhone or Android device:

Please click this URL to join.