(NEPTUNE) Situation Awareness and Smart Reconfiguration of Ad-hoc Military Electric Grids Using a Digital-twin
Abstract
Abstract:The proposed work aims to increase reliability and flexibility of military grids by providing 1)networked monitoring capabilities to identify, characterize, and report potential threats to thesystem security, and 2) an automated digital tool for adapting to network configuration changesthat re-optimize equipment use and set points. Our tools will make use of a ~digital twin~ that isan equivalent software version of the military grid. It will have a simple front-end for personnel tointeract with as a stand-alone application or as a plug-in to an existing application (e.g., ArcGIS).The project will connect the interface at the front-end with state-of-the art power systems solversand design new Artificial intelligence (AI) algorithms working in the back-end, including: 1)reinforcement learning algorithms to find the best configuration settings for generation assets(smart inverter, diesel generator) and controllable loads (HVAC, Electric Vehicles, storage, waterpumping, switches) prioritizing critical loads through the exploration of a variety of future possibleoperating conditions and, 2) secure machine learning techniques that provide status analysis,forecast future operating conditions and perform threat analysis using sensor feeds.The expected benefits for the Navy and DoD Benefit that come from the ~digital twin~ of theirelectrical network are as follows. The ~digital twin~ provides:1) a method to digitize and automatically update the system as the network, assets, and loadschange over time2) an interactive and accessible through an intuitive front-end interface,3) the capability to self-diagnosis issues and report concerns to Public Works using artificialintelligence:a. reinforcement learning to choose optimum policies and setpoints in response ofchanging operating conditions and objectives,b. machine learning to explore sensor outputs and provide situation awarenessregarding the grid status and forecasts of its possible future operating conditions.4) a secure inventory and activities log attained by using Blockchain to manage datacommunication between sensors and the digital twin.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Mar 11, 2020
- Source ID
- N000142012212
Entities
People
- Anna Scaglione
Organizations
- Arizona State University
- Office of Naval Research
- United States Navy