Ming Dong

PhD Student, Research Assistant,
Department of Electrical and Computer Engineering
University of Alberta
Edmonton, Alberta, Canada
T6G 2V4
Email: mdong@ualberta.ca


Biography

I received my B.Eng. degree in Electrical Engineering from Xian Jiaotong University, China in 2008. I am currently pursuing my Ph.D. degree with Electrical and Computer Engineering, University of Alberta, Canada. My research interests cover smart grid, power quality and power system protection.


Current Research

My research involves power system grounding, harmonic source modelling, power quality and its applications to smart grid and demand response. Currently, I am working on a project to extract dis-aggregated load information from smart meter data via power quality and other characteristics of loads.

Load disaggregation is a useful tool to extract detailed load information out of compound load data. It either dis-aggregates compound load into specific load level such as different appliances in a house or specific category of load level such as a lighting system in an office building. Through these techniques, users will be able to better understand their detailed energy uses and usage patterns of individual loads or load categories. Thus, decisions on how to save energy and reduce carbon footprint can be easily made. For the need of utility, understanding dis-aggregated load consumptions in an area is also key to help forecast future demand, develop demand response programs and design Time-Of-Use price to encourage load shifting and conservation.

Traditionally, there are two ways to obtain detailed load information: One is through userís behavior survey. The results collected may not be accurate since it is not possible for ordinary users to record their behaviors of load operations completely and correctly in the long term. Also, since power information of individual loads is unknown, it is not reliable to estimate energy consumption based on this; the other alternative approach is to install numerous energy measurement devices on each individual loads. However, it is expensive, inconvenient and hard for equipment maintenance.

My techniques are based on variety of load signatures such as active power, reactive power, harmonic contents, working cycle characteristics and other behavior characteristics. Utilizing above signatures, events of specific loads or load categories can be identified with only compound load data provided. Then, energy consumption of such loads and categories can also be calculated based on detected events. The approach is much more accurate than conducting survey and on the other hand, it does not require any effort or expense on extra equipment installations.

This project involves load signature studies, load identification algorithm for residential houses, load category estimation algorithm for commercial buildings and load signature extraction method studies. It aims to provide a solution kit to end-users of smart grid to help them make decisions on energy savings and demand response.