Doctoral Dissertations

Date of Award

8-2019

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Industrial Engineering

Major Professor

Xueping Li

Committee Members

Anahita Khojandi, Oleg Shylo, Jon Hathaway

Abstract

In this study, we aim to investigate the application of the green technologies (i.e., green roofs (GRs), Photovoltaic (PV) panels, and battery integrated PV systems) under climate change-related uncertainty through three separate, but inherently related studies, and utilize optimization methods to provide new solutions or improve the currently available methodsFirst, we develop a model to evaluate and optimize the joint placement of PV panels and GRs under climate change uncertainty. We consider the efficiency drop of PV panels due to heat, savings from GRs, and the interaction between them. We develop a two-stage stochastic programming model to optimally place PV panels and GRs under climate change uncertainty to maximize the overall profit. We calibrate the model and then conduct a case study on the City of Knoxville, TN.Second, we study the diffusion rate of the green technologies under different climate projections for the City of Knoxville through the integration of simulation and dynamic programming. We aim to investigate the diffusion rates for PV panels and/or GRs under climate change uncertainty in the City of Knoxville, TN. We further investigate the effect of different and evaluate their effects on the diffusion rate. We first present the agent based framework and the mathematical model behind it. Then, we study the effects of different policies on the results and rate of diffusion.Lastly, We aim to study a Lithium-ion battery load connected to a PV system to store the excess generated electricity throughout the day. The stored energy is then used when the PV system is not able to generate electricity due to a lack of direct solar radiation. This study is an attempt to minimize the cost of electricity bill for a medium sized household by maximizing the battery package utilization. We develop a Markov decision processes (MDP) model to capture the stochastic nature of the panels' output due to weather. Due to the minute reduction in the Li-ion battery capacity per day, we have to deal with an excessively large state space. Hence, we utilize reinforcement learning methods (i.e., Q-Learning) to find the optimal policy.

Comments

Each separate study has been or is being published in academic journals.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS