|Author:||Choi, Siu Leung|
|Title:||Study on power signature of home electric appliance for non-intrusive load monitoring (NILM)|
|Advisors:||Chen, Mingli (BSE)|
|Subject:||Electric machinery -- Monitoring|
Household appliances, Electric
Hong Kong Polytechnic University -- Dissertations
|Department:||Department of Building Services Engineering|
|Pages:||95 pages : color illustrations|
|Abstract:||Non-intrusive load monitoring (NILM) is one of the new power monitoring techniques. This technology can be integrated into home automation for the potential use of energy audit, occupancy detection and user activity detection in the residential or commercial space. The most possible application is the home energy management. It can provide the energy use information of each appliances namely real time and total energy usage, electric tariff and carbon footprint etc in the monitored area such as a house and office. The energy end users can improve their electricity usage by demand side management based on the feedback given by NILM. One problem of the NILM is that the accuracy of identifying the operated appliance may not be satisfactory. The power signature of electric appliance which is the unique electrical behavior of the appliance is a key factor to ensure NILM can identify the load. Selecting the suitable features of power signature is essential to enchance the identification accuracy. The objective of this study is to strengthen the accuracy or achieve the higher accuracy of appliance identification of NILM by selecting effective features of power signature. This study investigated the power signature of various home electric appliance in selected households by realizing their similarity and difference, so that some suitable features was estimated to distinguish the types of loads by comparing their power signature. Besides that, a power monitoring experiment with 10 cases was conducted to evaluate the identification accuracy of a wide range of the features. The result shows that shape features and multiple value features such as instantaneous current waveform, instantaneous power waveform and magnitude of 1st to 15th order harmonic current have significantly higher identification accuracy than the single value features such as real power, reactive power and root means square current etc. Two strategies including modified optimization algorithm for appliance identification and combining single value features into one feature can successfully improve the accuracy around two times of the original accuracy. In the optimal case, the combined single value features under the identification by the modified algorithms achieved 80% of accuracy almost as satisfactory as that of instantaneous current waveform. The potential of several single value features can be discovered to better the performance of NILM.|
|Rights:||All rights reserved|
Files in This Item:
|5725.pdf||For All Users (off-campus access for PolyU Staff & Students only)||1.61 MB||Adobe PDF||View/Open|
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item: