PMUs facilitate accurate monitoring of grid dynamics in real time. From the data obtained from PMUs, one can analyze sequence of events that have contributed to the catastrophic behavior of power system. A power system should work in a sinusoidal steady-state, characterized by a nominal frequency of either 50 or 60 Hz and thus the concept of phasor directly applies. The synchrophasor estimation is based on the same idea underlying the phasor, with the phase-angle calculated using Co-ordinated Universal Time (UTC) as a time reference. Therefore, it allows having a unique reference for all the sinusoidal signals to be measured in a wide or global area. Two synchrophasors, calculated at different points in a network, can be easily compared because they are related to common instants, because the time dissemination relies on satellite systems. Various research works are being carried out in analysis of PMU data.The fault in power system cannot be completely avoided. The longer it takes to identify and repair a fault, more harm may result in the electrical power system, especially in periods of peak loads, which could lead to the collapse of the system, causing the power outage to extend for a longer period and larger parts of the electrical network. It is essential to detect fault, its type and localise it. Incorporating PMU for this purpose is an area of ongoing research. The conventional methods used for detecting, classifying and localizing the fault are impedance measurement-based method and travelling wave method. PMUs are being increasingly used for detecting faults on transmission lines.
PMU are placed at optimal location in the transmission test system. A fault detection method is proposed using positive and zero sequence voltage measurements obtained from PMUs. This method can distinguish between faults and other disturbances/changes occurring in power systems. Once the fault is detected it is important to identify the type of fault. For this purpose a Support Vector Machine (SVM) has been proposed. The advantage of this method is that the input to SVM is measurements from PMU at generator bus alone. The structure of this paper is as follows. Section II explains the proposed fault detection method. The proposed SVM used for identifying the type of faults is explained in Section III. In Section IV, the proposed fault detection and classification algorithms are tested and validated on IEEE 14 bus system and results are discussed.