An Alarm and Fault Association Rule Extraction Method for Power Equipment Based on Explainable Decision Tree

For the widespread alarm logs in power generation and communication network equipment, it is difficult for professional maintenance staff to locate the alarm cause and remedy equipment faults. In terms of this issue, we propose an alarm-fault rule extraction method based on explainable artificial intelligence technology. We use an alarm time sequence diagram to express alarm information and build explainable alarm statistical features, and the important features are selected based on the weight random forest algorithm. Then we use the different combinations of important features to build a series of decision trees, which is understandable. We use this method to study the association rules between alarm and fault of synchronous digital hierarchy optical communication equipment in the power system. Experimental results show that this method is effective and reliable, thus provide an effective method for intelligent analysis of alarm logs and fault location.