Substation alarm information processing based on ontology theory

Logical alarm information is the key to mining the knowledge that hidden in the power system equipment operation and maintenance information. In this paper, a knowledge mining method of association rules based on ontology is proposed in order to realize the intelligent analysis of substation alarm information. Firstly, the alarm information ontology is based on the structure of substation configuration language, and it uses the methods of short text classification and data generalization, which lays the foundation for hierarchical database of the Apri ori algorithm. Secondly, this paper uses the ontology method to construct ontology class tree that integrates the unstructured records based on three dimensions of time, frequency and range. And then we traverse data, attributes, and relationships in the tree, so as to generate the data sources that computer can use for algorithms in data mining. Finally we use the improved Apriori algorithms based on alarm ontology to calculate and draw conclusions. Experimental results show that the efficient generation of association rules and assertions rules verify the validity of the method.