Publication Date

2017

Document Type

Thesis

Committee Members

Tanvi Banerjee (Committee Member), Yong Pei (Advisor), Mateen Rizki (Committee Member)

Degree Name

Master of Science in Computer Engineering (MSCE)

Abstract

The landscape of energy generation and utilization is witnessing an unprecedented change. We are at the threshold of a major shift in electricity generation from utilization of conventional sources of energy like coal to sustainable and renewable sources of energy like solar and wind. On the other hand, electricity consumption, especially in the field of transportation, due to advancements in the field of battery research and exponential technologies like vehicle telematics, is seeing a shift from carbon based to Lithium based fuel. Encouraged by 1. Decrease in the cost of Li – ion based batteries 2. Breakthroughs in battery chemistry research - resulting in increased drive range 3. Government incentives and tariff concessions by utilities for EV owners in the form of tax credits, EV – only parking spaces, free charging equipment etc., the automobile market, especially the passenger vehicle market, is witnessing a steady growth in the sale of electric vehicles. This has resulted in Electric Vehicles contributing to the electricity load resulting in two challenges 1. At the supply end, it contributes as a potential micro energy storage system to fit the time gap between the demand for electricity and the supply of renewable and/or low cost electricity generation; and, 2. At the consumer-end, it creates a necessity to make energy consumption as sustainable and renewable as possible, while preserving battery life. In this thesis work we attempt to provide multiple practical solutions to address these needs by advancing existing technologies in the industry. Firstly, we have developed a “Joint EV-Grid Solution for Robust and Low-Complexity Smart Charing”, where we have designed and implemented a distributed smart charging algorithm, which runs in the EV with load and pricing information collected from Grid through the charging station. It is responsible for optimizing the charge plan of the user’s vehicle based on his/her preference and ensure a full charge before departure. The objective could be minimizing the electricity cost per charge session or maximizing the renewable energy usage. For instance, by setting the preference to optimize the algorithm according to “Price”, the additional demand is scheduled to off-peak hours (i.e., incurring the least cost). Alternatively, by setting the preference to “Renewables” the EV charges based on the maximum availability of renewable energy sources, thereby maximizing the utilization of renewable energy resources which may lead to reduced cost, if not minimize it. Furthermore, we have improved on our initial approach by introducing “Smart Charging Solution through Usage/Charging Pattern Learning” where we have used machine learning algorithms like Logistic regression and Fuzzy Logic to enable EVs to learn the usage and charging pattern of users and prepare a charging plan that is personalized at the users’ end and prevents potential smart-changing-caused demand peaks by distributing the net load throughout the day. Through our experiment studies we were successful in creating a distributed Charging algorithm and a Machine Learning system that could cater to the said requirements through innovative charging strategies. Consequently, helping us create a sustainable, win – win situation for both electricity consumer and producer.

Page Count

114

Department or Program

Department of Computer Science and Engineering

Year Degree Awarded

2017

Download


Share

COinS