迴路負載監測系統及其非監督式方法之研究

能源的有效使用是智能電網的重要研究課題，而電網的負載監測技術則是進行此能源管理不可或缺的一部分。目前許多負載監測技術被開發出來，就監測技術的方式來區分，有直接量測目標負載資料的侵入式監測以及量測電力供應入口資料的非侵入式監測。其中辨識負載狀態是監測技術的重要項目，透過辨識的結果，系統能夠掌握各負載的狀態，依照需求自動調整能源使用的分配。此外，為了建立辨識時所需要的資料庫，一般有使用監督式方法與非監督式方法兩種方式。使用監督式方法需要事先取得訓練的樣本，但若系統中存在未知的負載資料，系統便無法進行辨識。為了提高非侵入式負載的辨識率，以及適應環境的變化，本研究提出非監督式方法應用在監測系統中，採取低解析度擷取、穩態特徵的方式，降低系統設置成本。使用K-means演算法分群，建立高斯混合模型(Gaussian mixture model, GMM)，代表各負載狀態性質；最大期望值算法(expectation-maximization algorithm, EM)估算模型參數。另外利用在線的資料提出適應性系統方法，不斷地在線資料對負載資料庫進行更新動作，並回饋迴路中負載監測資訊，進行系統的適應性微調。透過提出的適應失效指標，提供系統中的負載辨識區間與偵測未知負載的方法，賦予系統靈活性與延展性。實驗結果顯示，該方法有效地適應環境的變化及偵測出迴路未知新負載狀態的加入，在單一負載與並行負載辨識實驗中，較未使用適應性方法的83% 平均辨識率，有著更高的93% 以上的平均辨識率，另外在偵測迴路未知負載的實驗中，得到很好的偵測率。

關鍵字

非侵入式負載監測；智慧電網；高斯混合模型；非監督式；自適應系統；負載狀態辨識

並列摘要

Efficient energy use is an important research topic of the smart grid. Load monitoring is an integral part of energy management, convenient information, communication technology, and sensor applications. So far, many monitoring techniques have been developed, and non-intrusive load monitoring is one of them. However, there are many technical difficulties, which are still unable to achieve non-intrusive completely. The traditional methods by supervised learning required to obtain training sample. But if the unknown loads exist in the circuit, it will be causing much more difficult to monitor. In order to achieve the complete non-intrusive concept and to adapt to the changes in environment, this research proposes the unsupervised method that applied in the monitoring system, taking low frequency acquisition and steady-state feature extraction for reducing its setup costs. It uses the K-means clustering algorithm, building a Gaussian mixture model (GMM) to represent load status; the expectation-maximization (EM) algorithm is for estimating the model parameters. Additionally, the adaptive system is proposed. By getting real-time/online data, it is continuing to update the database and feedback the monitoring information of loads in the circuit. And it provides the identification interval and detection unknown loads methods by proposed failure detection index (FDI), which endows system with flexibility and scalability. The experiment shows that this method is effective to trace the changes in environment and to detect the unknown loads, making a complete solution in non-intrusive load monitoring system.

並列關鍵字

Non-intrusive appliance load monitoring ； Smart grid ； Gaussian mixture models ； Unsupervised ； Adaptive system ； NILM

參考文獻

[2] G. W. Hart, "Nonintrusive appliance load monitoring," Proceedings of the IEEE, vol. 80, pp. 1870-1891, 1992.

[7] S. McLaughlin, P. McDaniel, and W. Aiello, "Protecting consumer privacy from electric load monitoring," in Proceedings of the 18th ACM conference on Computer and communications security, 2011, pp. 87-98.

[8] A. Marchiori, "Enabling distributed building control with wireless sensor networks," in World of Wireless, Mobile and Multimedia Networks (WoWMoM), 2011 IEEE International Symposium on a, 2011, pp. 1-3.

[9] M.-C. Lin, Y.-C. Chen, S.-L. Tsao, and W. Tang, "Design and implementation of a home and building gateway with integration of nonintrusive load monitoring meters," in Industrial Technology (ICIT), 2012 IEEE International Conference on, 2012, pp. 148-153.

[10] L. Jiang, J. Li, S. Luo, S. West, and G. Platt, "Power Load Event Detection and Classification Based on Edge Symbol Analysis and Support Vector Machine," Applied Computational Intelligence and Soft Computing, vol. 2012, 2012.

國際替代計量

迴路負載監測系統及其非監督式方法之研究

全文下載

主題瀏覽