Honors Theses and Capstones
Date of Award
Spring 2022
Project Type
Senior Honors Thesis
College or School
CEPS
Department
Computer Science
Program or Major
Computer Science
Degree Name
Bachelor of Science
First Advisor
Marek Petrik
Abstract
Reinforcement learning (RL) techniques have been applied to smart grids with a variety of applications. The most common objective is to optimize profit for one actor in the system. The goal of this work is to apply different RL models to the smart grid at the Shoals Marine Laboratory (SML) located on Appledore Island, Maine in an effort to reduce costs and minimize the amount of nonrenewable energy consumed on the island. The RL models implemented resulted in more sustainable practices in simulations, with a linear spline model outperforming the naive policy the SML currently uses. Future work includes extending the RL model and more validation testing outside of the simulation.
Recommended Citation
Mattson, Daniel C., "Reinforcement Learning Applied to the Shoals Marine Laboratory Smart Grid" (2022). Honors Theses and Capstones. 658.
https://scholars.unh.edu/honors/658