Toward a Fair Transactive Energy Market: A Deeplearning Based Energy Consumption Prediction Model
Date of Award
Santa Clara : Santa Clara University, 2022.
Master of Science (MS)
Computer Science and Engineering
The application of machine learning is vast and quickly spreading across disciplines because of its versatile utility. By nature, machine learning implementations can quickly be obfuscated and ultimately introduce and perpetuate discriminatory practices, which leads to the issue of fairness. Transactive energy and the distribution of energy management technologies allow for new participants, meaning individual households and entities smaller than large energy providers, to enter the market to buy and sell energy. Machine learning has potential for meaningful use in many aspects of the transactive energy market process, and we focus on the specific aspect of how individual households can predict and manage their energy consumption to be a more competent and competitive agent in the market. This practice, aided by technologies like home energy monitors, can track the usage of specific appliances or outlets and dynamically change energy usage according to set or learned conditions. These devices and technologies can allow users to curate their energy consumption, but not all households can afford them or have access to them. If prediction models represent the majority of energy consumption patterns, then households with static consumption patterns could be unfairly represented, potentially causing them to pay higher energy prices than if they were fairly represented. By utilizing a reverse-engineered California Independent System Operator’s energy demand model, we may input our own energy data and patterns to find evidence of unfairness towards low-income households. This research focuses on the process of modeling energy consumption in California using machine learning techniques.
Hatori, Yuka, "Toward a Fair Transactive Energy Market: A Deeplearning Based Energy Consumption Prediction Model" (202). Computer Science and Engineering Master's Theses. 30.