Intelligent Data Mining Basedon Running Information Of Substations

With the rapid development of information technology and smart grid, mass data isgenerated in the daily operation of digital substation, resulting in the drastically expanding ofthe data sampled, accumulated, and to be analyzed in the system. Therefore it makes sense todiscuss how to get the hidden pattern and rules from large-scale, multi-dimensional data, inorder to provide decision support to operating personnel.This paper firstly introduces the current development of intelligent substation and thenanalyzes the fundamental principle, typical methods of data mining and application in powersystem substation automation technology. Summarizes the relevant theories of associationrule in data mining and propose an improved Apriori algorithm to dig out the association rulein the substation data. Original alarm data of the monitoring system classified and ruleassociated, we obtain device parameter information, status information, and relevanthistorical fault information. According to the rule association theory in data mining technique,extract the strong association rule that meets the minconfidence from mass data and dig outsystem’s operation status in each historical period, so as to provide basis of decision makingfor the safe operation of the digital substation.Taking background monitoring of substations in one district for example, based on itshistorical alarm data in daily operation, one automatic intelligent data mining system basedon substation operation data is constructed. Combined with the experience of the griddispatching personnel, this paper uses the above-mentioned algorithm to extract associatedrule from massive historical data, analyzes and corrects the extraction results, finally gets theexpert analysis rules which have practical guidance on real operation. The results show thatassociation rule extracting technique based on data mining can extract associated ruleinformation from mass data effectively. It is highly significant for operation personnel tograsp system status and analyze fault alarm timely and quickly.