Electronic Theses and Dissertations

Date

2020

Document Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Electrical & Computer Engineering

Committee Chair

Mohd. Hasan Ali

Committee Member

Eddie Jacobs

Committee Member

Muhammad Shah Jahan

Committee Member

John Hochstein

Abstract

A smart building is the one that is equipped with automated systems such as lighting, shading, heating, ventilation, and air-conditioning (HVAC), etc., for fulfilling consumers demand. An efficient load forecasting system helps the building energy management system (BEMS) schedule the loads, operate the energy sources and energy storage systems effectively during peak hours to reduce cost of energy and remove burden on the grids. Furthermore, the load scheduling is a key element for demand side management (DSM) system to actively participate in demand response program. Researchers have been investigating on improved and effective load forecasting and load scheduling methods over the last decade. The conventional methods such as the artificial neural network (ANN) technique needs a lot of previous or historical data for training and learning. Moreover, the correlation between the inputs and output is very crucial for better performance of the ANN methods. Similarly, other conventional methods such as random forest, LSBoosting and long short-term memory (LSTM) have their own drawbacks.In order to overcome the drawbacks and limitations of conventional methods, this dissertation proposes new methods for load forecasting and scheduling. An effective real-time health monitoring system with a view to checking the health condition of all loads and getting a pre-alert before the advent of a disaster or faults is proposed. In addition, a web-based application to be accessed by any smart device such as smart television, mobile phone, etc., is suggested. Moreover, the impact of increased energy consumption during office hours due to COVID-19 pandemic on the local distribution transformer is analyzed. Also, the behind-the-meter (BTM) sources to mitigate the adverse effect of increased load on distribution transformer is proposed.Simulation results performed by MATLAB/Simulink software indicate that the proposed load forecasting and scheduling methods perform better than conventional methods. The proposed health monitoring system is effective and can provide both normal operating and alert messages based on the operating conditions. Moreover, it is found that the load consumption has increased due to the COVID-19 lockdown situations. Finally, the proposed BTM solution can mitigate the adverse effect of increased loads and reduces the transformer loss.

Comments

Data is provided by the student.

Library Comment

Dissertation or thesis originally submitted to ProQuest

Download

Share

COinS