The Research On Artificial Fault Diagnosis System Of Natural Gas Compressor

In CNG gas stations, the compressor would burn and explode once compressible medium which was flammable and explosive gas reached limiting concentration. The running state of compressor which was normal or not would influence the economic benefit of gas station directly, e ven cause safety accidents. The advanced technology like data acquisition, automatic control, computer network technology, artificial intelligence, etc were used synthetically to develop real time monitoring and intelligent fault diagnosis system which became developing trend for natural gas compressor. Local monitor for single natural gas compressor would extend to remote monitor for more than one equipment to realize centralized management of device and improve real monitoring, fault diagnosis efficiency of natural gas compressor.The natural compressor from Xi’an Kunlun Hydraulically-driven Machinery Factory was the object of study. The comprehensive research of natural gas compressor intelligent fault diagnosis system was caught out combining with the actual situation and compressor client needs.Firstly, the Operating principle was elaborated simply, compressor common fault which laid the foundation for the case base. in the running processes was summarized. The natural compressor intelligent fault diagnosis system was established integrating Rule-Based Reasoning and Case-Based Reasoning based on the equipment fault diagnosis technology at home and abroad.Secondly, in this paper, on the basis of comprehensive analysis for knowledge acquisition, knowledge representation, retrieval methods, system modification of fault diagnosis system, the integrated reasoning model combining with RBR and CBR was established: factitious seek ways actively was used to get knowledge from domain experts, the method of decision tree was used to show regulations, case framework described the fault case, rule library and case library was integrated by database, derivative analogy method and k-Nearest Neighbor was used to modify the case and search the system. The whole operation principle of natural compressor intelligent fault d iagnosis was completed. At last, Java programming language was used to develop intelligent fault diagnosis system which realized the real monitoring and intelligent fault diagnosis intensively for natural compressor fault based on MySQL5.5 database. The natural compressor in gas station could be monitored in real time to analyze the fault reasons and position automatically, quickly and accurately. The intelligent fault diagnosis system could provide maintenance plan for serviceman to solve the bottleneck problem of natural gas compressor fault diagnosis effectively.Feasibility, veracity and flexibility of developing intelligent fault diagnosis system were verified in spot test of enterprise in this paper. The system would reduce the labor intensity of serviceman effectively, decrease the additional costs on business and equipment maintenance, lower the economic losses resulting from shutting down of natural compressor. it has a good practicability for enterprise and user.