Title:  Improving Distribution System Model Accuracy by Leveraging Ubiquitous Sensors

Committee:

Dr. Santiago Grijalva, ECE, Chair , Advisor

Dr. Ronald Harley, ECE

Dr. Maryam Saeedifard, ECE

Dr. Sakis Meliopoulos, ECE

Dr. Shabbir Ahmed, ISyE

Abstract: 

This dissertation addresses the need for utilities to improve the analytical and operational distribution system modeling accuracy as well as to manage the Big Data from modern distribution system measurement sources for future advanced distribution automation schemes with ubiquitous distributed energy resources. In particular, this dissertation presents accurate, flexible, and computationally efficient parameter and topology estimation methods to calibrate existing utility distribution system secondary circuit low-voltage models. Several methods are presented to handle unknown and known secondary circuit topologies and different available measurement datasets. The presented methods leverage the Big Data generated by the modern distribution system measurements from AMI and PV micro inverters. This dissertation also presents data validation and imputation methods to manage the granularity, accuracy, and reliability issues related to the modern distribution system measurements. The developed data validation methods were effectively used to detect numerous issues in Georgia Tech AMI. Compared to conventional approaches, the developed computationally efficient data imputation method has a superior average accuracy in imputing Georgia Tech AMI measurements. The method creates a series of imputed samples that have a continuous profile with respect to the adjacent available measurements, which is a highly desirable feature for time-series analyses.