Date of Award

Summer 2020

Degree Type

Thesis

Degree Name

Master of Science (MS)

Department

Electrical Engineering

Committee Chair

Josh Wold

First Advisor

Matt Donnelly

Second Advisor

Dan Trudnowski

Third Advisor

Curtis Link

Abstract

The power industry requires the validation and sharing of the models for generators and controlling equipment to other utilities within North America. Accurate simulation of the electric grid is essential when setting operation and control strategies during various conditions present on the grid. An industry accepted method for performing model validation involves using the voltage and frequency measurements during an event, typically recorded by a phasor measurement unit, to perform a dynamic simulation on the system. The output real and reactive powers from the dynamic simulation and the actual event are then compared to each other. If the real and reactive powers closely match, the overall model is considered validated. Otherwise, the process of model calibration is performed to make improvements to the model. Least-squares fitting is an industry recognized method used during model calibration. No measurement of uncertainty is provided for individual parameters when sharing models between generator owners, utilities, and regulatory bodies. This thesis proposes a methodology for finding a measurement of uncertainty for each model parameter in the form of the standard deviation of the parameter value by means adding measurement noise and least-squares fitting.

Comments

A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Electrical Engineering

Share

COinS