CRII: OAC: A (near) Real-time Framework for Smart Integration of Electric Vehicles to Microgrids

University Of Florida, Gainesville FL

Investigators

Abstract

This award is funded in whole or in part under the American Rescue Plan Act of 2021 (Public Law 117-2). The growth of electric vehicle (EV) adoption creates a timely opportunity for utilities and power systems to boost revenue and build sustainable load growth. However, uncontrolled EV to power grid integration, particularly in smaller-scale power systems such as microgrids, brings significant problems, e.g., power flow fluctuation and unacceptable load peaks reducing power network reliability and power quality. In this project we develop a framework to use the flexible energy capacity of EVs to provide services for microgrids. This method increases EV owners’ engagements and will enhance social welfare for all shareholders of the vehicle to microgrid (V2M) connection including EV owners, intelligent charging stations (ICSs), and microgrids. Considering the vast amount of data required for managing a large number of EVs in a microgrid, the project will employ advanced machine learning techniques and analytical approaches to reduce the computational complexity of the problem. The proposed framework will provide an understanding of the characteristics and requirements for future cyberinfrastructure design, particularly in integrated large-scale ecosystems. This work represents a broad, novel contribution to literature, education, outreach, and diversity; as such, the project aligns with NSF’s mission to promote the progress of science and to advance prosperity and welfare. The overall aim of the project is to develop a (near) real-time low computational cost framework for V2M integration. The research incorporates concepts of load prediction and game theory advancing V2M technology to use the flexible energy capacity of EVs in microgrids according to the needs of both. Traditional clustering and forecasting methods cannot be directly applied to microgrids due to their high volatility load profile caused by coupling various distributed energy resources. The technical contribution of the project is threefold: 1) to design an innovative machine learning-enabled algorithm for short-term load forecasting in microgrids that consists of robust data preprocessing, feature extraction, and a selection algorithm, followed by a weighted advanced long short-term memory (WA-LSTM) model. The selected features have high relevance and minimum redundancy reducing the computational cost significantly for the proposed WA-LSTM to predict the load profile; 2) to develop a cooperative game to capture interactions among EVs and their corresponding ICSs that outlines (dis)charging profiles for EVs considering vehicle parameters and constraints. A Nash bargaining game with relaxed constraints and a new penalty distribution policy is proposed to maximize the profits of all players in the coalition while satisfying the microgrids' requirements as the global goal; 3) to evaluate the proposed methods in experimental testbeds using hardware in the loop (HIL) setups to provide both analytic and experimental evidence to demonstrate the effectiveness of the proposed solutions. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.