Decomposition Techniques for Power System Analysis

  • Author / Creator
  • In recent years, the increased public awareness of energy conservation has attracted serious attention to detailed energy consumption monitoring and management for the end users of power system. Load decomposition is a technique that can extract detailed sub-load information from compound load information. This technique decomposes a compound load such as an entire residential house into specific sub-load levels such as different home appliances by using only the aggregated metering data of the compound load. Through load decomposition, users can better understand the usage patterns of individual loads or load groups and therefore decide on how to save energy. On the utility side, load decomposition can be very helpful for load forecast, demand response program development, and Time-Of-Use price design. In the past, traditional methods are either too costly or inaccurate. Therefore, some researchers proposed a non-intrusive load monitoring (NILM) approach that can identify and track major sub-loads based on only the total signal collected from the meter-side with acceptable error. Recently, the vast deployment of smart meters has raised considerable interests in this approach. However, many critical problems still need to be solved before it truly becomes technically available for ordinary end-users. To solve the above problems, at the beginning, this thesis presents a novel NILM method based on event detection and load signature studies. The key idea is to model the entire operating cycle of a load and make identification based on event-window candidates. The proposed technique makes NILM more applicable for complex loads, more robust for load inventory change and can also simplify the training process; on the other hand, the thesis addresses a new and critical problem that previous researchers ignored---the non-intrusive extraction of load signatures. The proposed approach is an unsupervised non-intrusive approach which can automatically extract load signatures by using the meter-side data and requires almost zero effort from users. This thesis also discusses how to estimate the energy for several key components in a residential house such as the NILM identified appliances, load groups and background power. Based on estimation, residential energy characteristics are discussed with respect to the Time-of-use price.

  • Subjects / Keywords
  • Graduation date
    Fall 2013
  • Type of Item
  • Degree
    Doctor of Philosophy
  • DOI
  • License
    This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.
  • Language
  • Institution
    University of Alberta
  • Degree level
  • Department
  • Specialization
    • Energy Systems
  • Supervisor / co-supervisor and their department(s)
  • Examining committee members and their departments
    • Xu, Wilsun (Electrical and Computer Engineering)
    • Dick Scott(Electrical and Computer Engineering)
    • Ardakani Masoud (Electrical and Computer Engineering)
    • Musilek Petr(Electrical and Computer Engineering)
    • Dinavahi Venkata (Electrical and Computer Engineering)
    • Bhattacharya,Kankar (Electrical and Computer Engineering, University of Waterloo)