Application of Wavelet Entropy and RBF Technique Based on Initial Current Travelling Wave for Transmission Line Protection

Abstract

The ability to accurately detect and classify the type of faults as well as locate the fault distance plays a great role in the protection of power transmission system. This paper presents a technique that processes the initial current traveling wave by wavelet packet transform (WPT) to yield the wavelet entropy at appropriate level of decomposition. The extracted features are applied to radial basis function neural network (RBFNN) for decision of fault or non-fault at any phase of the transmissions line. Once the faulty phase has been identified, the fault location away from the relay can be accurately estimated by using other RBFNN correspondingly. The algorithm employs wavelet packet transform and artificial neural network to improve the performance associated with conventional voltage or current traveling-wave-based schemes due to effect of factors such as fault inception angle, fault distance, insufficient high-frequency decomposition and uncertain principle threshold values. The proposed technique is tested with wide range of operating conditions and provides accurate results for fault classification and location, respectively.