2026-05-24T11:13:21Zhttps://keep.lib.asu.edu/oai/requestoai:keep.lib.asu.edu:node-1579192024-12-20T18:25:12Zoai_pmh:alloai_pmh:repo_items157919
https://hdl.handle.net/2286/R.I.55621
http://rightsstatements.org/vocab/InC/1.0/
2019
92 pages
Masters Thesis
Academic theses
Text
eng
Hashmy, Syed Muhammad Yousaf
Weng, Yang
Sen, Arunabha
Qin, Jiangchao
Arizona State University
Masters Thesis Electrical Engineering 2019
Due to the rapid penetration of solar power systems in residential areas, there has<br/><br/>been a dramatic increase in bidirectional power flow. Such a phenomenon of bidirectional<br/><br/>power flow creates a need to know where Photovoltaic (PV) systems are<br/><br/>located, what their quantity is, and how much they generate. However, significant<br/><br/>challenges exist for accurate solar panel detection, capacity quantification,<br/><br/>and generation estimation by employing existing methods, because of the limited<br/><br/>labeled ground truth and relatively poor performance for direct supervised learning.<br/><br/>To mitigate these issue, this thesis revolutionizes key learning concepts to (1)<br/><br/>largely increase the volume of training data set and expand the labelled data set by<br/><br/>creating highly realistic solar panel images, (2) boost detection and quantification<br/><br/>learning through physical knowledge and (3) greatly enhance the generation estimation<br/><br/>capability by utilizing effective features and neighboring generation patterns.<br/><br/>These techniques not only reshape the machine learning methods in the GIS<br/><br/>domain but also provides a highly accurate solution to gain a better understanding<br/><br/>of distribution networks with high PV penetration. The numerical<br/><br/>validation and performance evaluation establishes the high accuracy and scalability<br/><br/>of the proposed methodologies on the existing solar power systems in the<br/><br/>Southwest region of the United States of America. The distribution and transmission<br/><br/>networks both have primitive control methodologies, but now is the high time<br/><br/>to work out intelligent control schemes based on reinforcement learning and show<br/><br/>that they can not only perform well but also have the ability to adapt to the changing<br/><br/>environments. This thesis proposes a sequence task-based learning method to<br/><br/>create an agent that can learn to come up with the best action set that can overcome<br/><br/>the issues of transient over-voltage.
Electrical Engineering
PV System Information Enhancement and Better Control of Power Systems.