This project will use intelligent building load management to improve energy resilience of the base. In particular, this project will:

• Develop digital twin of a selected military microgrid.
• Demonstrate the ability to quantify building load flexibility using advanced building load modeling, voltage to power sensitivity analysis, and site-specific weather forecasting methods.
• Demonstrate the ability to quantify the impacts of load disturbances on microgrid voltage quality and system stability.
• Demonstrate, through hardware-in-the-loop experiments on the digital twin, advanced load management by using multi-agent deep reinforcement learning (MADRL) proactive control algorithms to manage building loads and other onsite distributed energy resources for (1) improving energy efficiency and system power quality, and (2) minimizing the amount of load shedding while maintaining system stability during grid outages.
• Hold facilitated meetings with installation key stakeholders including mission owners to collect inputs on if and how the developed approaches can be applied in the field.

This project will use site-specific weather forecasting and model-predictive control methods to quantify building load flexibility, and use global sensitivity analysis to analyze the impact of load disturbances on microgrid voltage quality under grid-tied operations and system stability under islanded operations. With building load flexibility and load disturbance impact quantified, MADRL algorithm will be implemented to improve energy efficiency for normal gridconnected operations and enhance energy resilience for islanded operations under grid outages. All technologies will be demonstrated on the digital twin system that will be created for Marine Corps Air Station Miramar microgrid. The experiment results will be used to inform real system operations in the base.

The outcome of this project will enable intelligent single-building loads management and multi-building coordination. The technology provides cost-effective solutions for the Department of Defense facilities to:

• Minimize load shedding by optimally pre-heating/cooling the building and operating photovoltaic and batteries, and increase the capability of the base for providing energy supply for critical loads by at least 20%.
• Be aware of impacts of load disturbances on energy system voltage quality and stability, and proactively mitigate these impacts by harnessing load flexibility and intelligent control of various resources and maintain stable islanded microgrid operations.
• Manage critical infrastructures for energy efficiency and resilience more effectively rather than depending on costly new installations.
• Reduce peak load and achieve non-intrusive and sustainable energy savings for building energy systems and base microgrids with at least 5% of annual energy savings.