電力系統中異常事件的偵測、分析和特徵化之研究

隨著科技的進步，人民對用電的品質需求愈來愈高，電力系統發生擾動將使得電壓、電流偏離額定值，可能導致用戶端的設備嚴重損害，並造成損失，因此輸電線路的故障預測，是電力系統中重要的研究項目之一。在電力系統發生故障之前，如果能快速的診斷出故障種類和位置，則電力公司可以免於突發性故障，並節省人力上的調動、減少停電時間以及停電時間上的損失。電力系統中的各種設備會因為電力來源的不穩定而影響設備的運作和生產的品質，甚至增加公司的虧損。為了在嚴重的電力故障事件發生以前，提前得知故障的發生，必定能減少許多損失，因此，蒐集各種異常波形的種類，找出偵測電力異常之方法，加以分類和分析，最終達到預測故障徵兆的發生。本論文利用統計中的機率密度函數得出各個波形的高斯分佈圖，加上Kullback-Leibler divergence (KLD) 方法來得出異常事件與正常事件散佈圖的差異度，進而存取異常事件，取事件之特徵值後進入最近鄰居分類法和機器學習技術中的支撐向量機統整所有異常事件，比較兩個分類器的效能差別後，完成故障預測系統，從而提供快速、準確的故障判斷。

關鍵字

故障預測；高斯分佈； KL散度；最近鄰居分類法；支撐向量機

並列摘要

In recent years, the widespread use of power quality (PQ) monitors and research advancements in a new research field named power quality data analytics. Power quality disturbance data is increasingly applied to extract useful information about the conditions of power system, such as monitoring incipient equipment failures, and solve various power system problems based on the information. In this thesis, abnormalities are detected by comparing with and without disturbances. Kullback-Leibler divergence (KLD) is used to assess the difference of the distributions. An abnormality exists if the KLD is larger than a threshold. The KLD could be used as features. To precisely forecast the incipient fault event, k-nearest neighbors (KNN) and support vector machine (SVM) are used to classify different abnormal waveforms with features of numerous abnormal events. With extend of characteristic of KLD, a diagnosis method is proposed. Results show that the system can provide fast and accurate fault forecast in the power system.

並列關鍵字

Fault Forecast ； Gaussian Distribution ； Kullback-Leibler divergence ； k-nearest neighbors ； support vector machine

參考文獻

[1] L. Benzhe, J. Yindi, and X. Wilsun, “A Generic Waveform Abnormality Detection for Utility Equipment Condition Monitoring,” IEEE PES Transactions on Power Delivery, Jane 2016.

[2] E. Byvatov, U. Fechner, J. Sadowski, and G. Schneider, “Comparison of Support Vector Machine and Artificial Neural Network Systems for Drug/Nondrug Classification,” Journal of Chemical Information and Computer Sciences, vol. 43, Sep. 2003, pp. 1882-1889.

[3] Tarlochan S. Sidhu and Zhihan Xu, “Detection of incipient faults in distribution underground cables,” IEEE Trans. Power Del. Vol. 25, no. 3, pp. 1363-1371, Jul. 2010.

[5] J. A. Wischkaemper, C. L. Bernner, B. D. Russell and K. Manicannan, “Application of waveform analytics for improved situation awareness of electric deistribution feeders,” IEEE Trans. Smart Grid, vol. 6, no. 4, pp.2041-2049, Apr. 2015.

[7] S. Bhowmick and S. Namdi, “Online detection of an intern winding fault in single-phase distribution transformers using a terminal measurement-based modeling technique,” IEEE Trans. Power Del. Vol. 30, no. 2, pp. 1007-1015, Apr. 2015.

國際替代計量

電力系統中異常事件的偵測、分析和特徵化之研究

未授權

主題瀏覽