A Hybrid Approach for Big Data Outlier Detection from Electric Power SCADA System

摘要: SCADA (Supervisory Control and Data Acquisition) databases have three main features that identify them as big data systems: volume, variety velocity. SCADAs are extremely important for the safety security operation of modern power system provide essential online information about state to operators. A current research challenge is efficiently process this data, which involves real-time measurements hundreds thousands heterogeneous electrical physical measurements. Among foreseen automation tasks, outlier detection one most mining techniques systems. However, like others techniques, traditional fails when dealing with problems in volume dimensionality high ones observed a SCADA. This work aims at circumventing these restrictions by presenting methodology consists pre-processing algorithm hybrid approach detectors. The assessed using real from Brazilian utility company. results show proposed capable identifying outliers correlated events affect system.