Study Of An On-line Intelligent Alarm

With the continuous expansion of the scale of power grids, various amount of information which gathered together the power grid control centre surges, make dispatcher for change quickly of alarm information indecisive, drown in the sea the quantity of alarm information, Make alarm information can not play its due role. Data mining(DM) is a process that can generate a credible, innovative, effective and can be understood model from a large number of historical data. DM is to discover regularity and the trend of the problem which are hided behind large amounts of data, in the use of real-time alarm information with the established rule,it can reflect the characteristics of information,remove the information of noise pollution and supply the power net condition for the operator.this paper,which includs data mining knowledge ,using the decision tree based on the algorithm constructs a SCADA alarm information to identify the information.we can improve the alarm information after removing the noise data,and better by going through the inference engine,then it is releasd to the dispatcher in the text or more acceptable manner.as a reasult,with the help of the alarm information the decision making supporting system can become more intellectual.