以基因演算法為基礎之非侵入式負載監測系統設計

A Genetic Algorithm Based Non-intrusive Load Monitoring System Design

蕭玉庭(Yu-Ting Hsiao)

指導教授 ： 楊宏澤

中原大學/工學院/電機工程研究所/碩士(2001年)

https://doi.org/10.6840/cycu200100310

摘要

傳統的負載監測系統監測負載的電力使用情形時，通常需要加裝硬體設備在所要監測的負載上，而非侵入式負載監測系統只需安裝監測器在電力入口處，藉由分析入口處的電力波形就能得知負載的電力使用情形。由此可知非侵入式負載監測系統較傳統的侵入式負載監測系統所需的監測設備與成本為少且較容易安裝、移除及維護，但其所需的負載辨識技術卻較侵入式負載監測系統來的複雜。
目前非侵入式負載監測系統主要是以實功及虛功兩特徵值來做負載辨識，為探討其他有效之特徵值，本文以適應性基因演算法配合圖樣識別技術建立一個特徵搜尋與負載監測辨識系統，並分別以實測資料及電磁暫態程式模擬訊號兩個案例來探討不同性質的特徵向量，藉此瞭解各種不同性質之特徵向量的辨識成效。本文透過兩案例的研究，驗證所建立之非侵入式負載監測系統的性能。
從兩案例的結果探討中，本文發現所建構之系統可驗證實功及虛功為目前最有效的負載辨識特徵，並進一步發掘其它如電流諧波量特徵值與利用小波轉換暫態波形等方法，有助於負載辨識的工作。此外，系統亦組合不同性質之特徵值中具有較佳辨識能力的特徵值，並給予特徵值適當的權值以強化系統的負載辨識能力。本文所開發適應性基因演算法結合圖樣識別技術建搆之方法初步可作為非侵入式負載監測系統之用。

關鍵字

負載辨識 ； 基因演算法 ； 圖樣識別 ； 非侵入式負載監測系統

並列摘要

Observing the energy consumption of electric loads, convectional load monitoring system needs to install hardware circuit on each load to be monitored. However, non-intrusive load monitoring system (NILMS) only needs to install a monitoring device on the electric power entrance point to collect the data for energy consumption of the loads by analyzing the signal waveforms collected and identifying the loads accordingly. Therefore, the monitoring facilities and cost needed by the NILMS is less than those of the conventional one. In addition, NILMS is also easier to install, remove and maintain. But the techniques needed for load identification are more sophisticated as compared with the convectional one.
The load identification in the NILMS mainly depends on the features of real and active powers. To investigate the efficiency of the other features that can be used for load identification, The thesis builds a feature-searching and load-monitoring identification system by integrating adaptive genetic algorithm with pattern recognition techniques. Based on the measured signals and the simulated data by using the electromagnetic transient program, different combinations of feature vectors are examined to improve the accuracy of the load identification. Two case studies are employed to verify the performance of the NILMS developed.
From the results obtained, it is found that the real and active powers are verified to be the most efficient features for the load identification. In addition, the features of current harmonics and the wavelet transformed coefficients of the transient signals are found to be helpful for the load identification. Moreover, this system combines and gives suitable weights to the features, which have superior performance of characterizing the loads, to enhance the load identification ability of the system. The methodologies developed by using adaptive genetic algorithm combined with pattern recognition techniques are suitable for a preliminary NILMS.

並列關鍵字

load identification ； non-intrusive load monitoring system ； pattern recognition ； genetic algorithm

參考文獻

[1]S. Drenker and A. Kader, “Nonintrusive Monitoring of Electric Loads,” IEEE Computer Applications in Power, vol. 12, pp. 47-51, 1999.

[2]G. W. Hart, “Nonintrusive Appliance Load Monitoring,” Proceedings of the IEEE, vol. 80, no. 12, pp. 1870-1891, December 1992.

[3]G. W. Hart, “Residential Energy Monitoring and Computerized Surveillance via Utility Power Flows,” IEEE Technology and Society Magazine, vol. 8, pp. 12-16, 1989.

[4]A. Cole and A. Albicki, “Nonintrusive Identification of Electrical Loads in a Three-Phase Environment Based on Harmonic Content,” IEEE Conf. on Instrumentation and Measurement Technology, vol. 1, pp. 24-29, 2000.

[5]A. I. Cole and A. Albicki, “Data Extraction for Effective Non-Intrusive Identification of Residential Power Loads,” IEEE Conf. on Instrumentation and Measurement Technology, vol. 2, pp. 812-815, 1998.