Online Applications of Wavelet Transforms to Power System Relaying - Part II

Recent wavelet developments in power engineering applications, include detection, localization, classification, identification, storage, compression, and network/system analysis of the power quality disturbance signals, to very recently, power system relaying [1,2,3,4]. This paper assesses the online use of wavelet analysis to power system relaying. The paper presents a novel technique for transmission-line fault detection and classification using the DWT for which an optimal selection of mother wavelet and data window size based on the minimum entropy criterion has been performed. The paper starts with the review of recent work within the field of wavelet analysis and its applications to power systems engineering. Then, the theoretical background of the technique is presented and the proposed method is described in detail. Finally, the effect of different parameters on the algorithm are examined in order to highlight its performance. Typical fault conditions on a practical 220 kV power system as generated by ATP/EMTP is analyzed with Daubechies wavelets. The performance of the fault classifier is tested using MATLAB software. The feasibility of using wavelet analysis to detect and classify faults is investigated. Finally it discusses the results, limitations and possible improvement. It is found that the use of wavelet transforms together with an effective classification procedure is considered to be straightforward, fast, computationally efficient and allow for real-time accurate applications in monitoring and classifying techniques in power engineering.