Building The Knowledge Base Of The Mine Ventilators’ Fault Diagnosis Based On FTA

Now, most companies have all kinds of rotating machinery condition monitoringsystem, which need on-site technical personnel to deal the further analysis and diagnosiswith real-time monitoring data. But, at present, the site operation personnel of miningenterprises have a common problem that the relevant professional level they have islimited. It is difficult for them to do the subsequent analysis and diagnosis with real-timemonitoring data. Therefore, the fault diagnosis expert system can solve the contradictionbetween the requirements of mining enterprises and the low level of site operation staffswell. For that, the construction of knowledge base is focused on in this paper which is thecore part of the mine fan fault diagnosis expert system.The research object is the mine fan in this paper. The common failure mode, failuremechanism, fault symptoms and cause of fault was analyzed and summarized. On thebasis of these methods above, a method to establish a knowledge base basing on the faulttree was proposed; and it proposed a train of thought of the establishment of the fault treewhich made fault features as the top event.The mine fan fault tree was established, whichbased on characteristic frequency including the fundamental frequency characteristics,fractional frequency characteristics, double frequency characteristics and the highfrequency characteristics.It built the mine fan structure of knowledge base which was made of the faultframework, framework rules, fault rules, fault symptoms, the cause of the problem andmeasures. It completed the knowledge storage with the library, rule base andcountermeasures for the library. This paper was a blend of the concept of fault tree andthe framework of knowledge representation, the minimum cut set of fault tree and therules of knowledge representation, the end event of the minimum cut set and conditionsof rules, and completed the construction of the mine fan fault diagnosis expert systemknowledge base with SQL database language.At last, it analyzed the probability of top event, probability of bottom events, theimportance of bottom events probability, and made a instance analysis of doublefrequency signal with the knowledge in the knowledge base which was constructed bythe fault tree. It verified the rationality and effectiveness of the knowledge base basing on the fault tree.