NEURAL NETWORK BASED TRANSFORMER DIFFERENTIAL PROTECTION
Abstract: The role of a differential relay for power transformer is to trip during fault condition and blocks the tripping during inrush, overexcitation, CT saturation conditions of the power transformer. Conventional harmonic restrained relay may mal operate due to the presence of the second and fifth harmonic during internal faults because of non-linear loads and capacitance in the transmission lines. This project presents a technique for classifying transient phenomena in power transformers for differential protection. Discrimination among different operating conditions (i.e., normal, inrush, overexcitation, CT saturation) and internal faults of the power transformer is achieved by differential relay using the algorithm based on wavelet transform with Neural Network. The wavelet transform is a powerful tool in the analysis of the power transformer transient phenomena because of its ability to extract information from the transient signals simultaneously in both the time and frequency domain. Neural network is used because of its self-learning and highly nonlinear mapping capability. Keywords – Wavelet Transform, back propagation neural netwok, transformer differential protection. I.Introduction Differential protection is the primary protection for larger power transformers. It contains the operation by all types of internal faults, and blocks the operations of the transformer by inrush, overexcitation and external faults. Since a magnetizing inrush current generally contains a large second harmonic component in comparison to an internal fault, conventional transformer protection systems are designed to restrain during inrush transient phenomenon by sensing this large second harmonic. However, the second harmonic component may also be generated during internal faults in the power transformer, due to CT saturation or the presence of a shunt capacitor or the distributive capacitance in a long EHV transmission line to which the transformer may be connected. Moreover, the second harmonic components in the magnetizing inrush currents tend to be relatively small in modern large power transformers because of improvements in the power transformer core material (Hi-B and D electrical steel) . Consequently, the commonly employed conventional differential protection technique based on the second harmonic restraint, will thus have difficulty in distinguishing between an internal fault and an inrush current thereby threatening transformer stability. Alternative, improved protection techniques for accurately and efficiently discriminating between internal faults and inrush currents have thus to be found. Recently, Artificial Neural Network (ANN) techniques have been applied to singlephase and three-phase power transformer protection to distinguish internal faults from magnetizing inrush currents -. However, the ANNs in these existing studies are specific to particular transformer systems, and would have to be retrained again for other systems. Moreover, the employed feature extraction techniques are based on either time or frequency domain signals, and not both time and frequency features of the signal; this is very important for accurately distinguishing between an internal fault and inrush current. The wavelet transform is a relatively new and powerful tool in the analysis of the power transformer transient phenomenon because of its ability to extract information from the transient signals simultaneously in the time and frequency domain, rather than conventional Fourier Transform, which can only give the information in the frequency domain. Recently, the wavelet transforms have been applied to analyze the power system transients, power quality, as well as fault location and detection problems . This paper presents an algorithm for the differential relay to distinguish between internal faults and magnetizing inrush, overexcitation, CT saturation currents. The...
Please join StudyMode to read the full document