Network Monitoring-based Automatic Alarm Against Operational Abnormal States of Substations
Automatic alarm is a key function of substation automatic monitoring system,but currently prevailing methods are inferior in accuracy due to high false negative ratios and false alarm ratios.The present work made a preliminary at-tempt to establish a network monitoring-based method for the automatic alarm of operational abnormal states.The method divides the substation into multiple heterogeneous monitoring areas through multi cluster partitioning,uses wire-less network to transmit abnormal variable data,utilizes RVM algorithm for data classification and abnormal feature ex-traction,and identify and alarms against operational abnormal states through reliability model.The method was experi-mentally indicated to achieve a false negative ratio and a false alarm ratio both below 1%,and hence to have a potential applicative prospect in the field of substation automatic monitoring.
network monitoringsubstationabnormal stateautomatic alarmRVM algorithmreliability model