ABSTRACT
With sustained development of national economy, people’s demand for energy becomes more and more strongly, but the shortage of oil and natural gas causes the contradiction between supply and demand of the energy. Now coal-bed methane extraction has alleviated the problem of energy shortage partly. The project of coal-bed methane gathering conducting in Panhe where is about 80 kms northwest of jincheng, Shanxi Province, and is under the jurisdiction of Qinshui. 200 new CBM wells are exploited in this project. The CBM gathering system consists of the well site, gas pipe network, gathering station and centralized processing booster station to complete CBM’s gathering, purification, separation, pressure regulation and transmission[1]. In order to improve the utilization of coal gas and ensure fault diagnosis of the key parts timely and effectively, the fault diagnosis mainly focuses on single wellhead[2] and booster station compressor. The compressor fault diagnosis based on Fuzzy Neural Network Associative Memory Model not only has the ability of expressing fuzzy and qualitative knowledge, but also has self-learning and self-adaptive ability as well as nonlinear expression ability. In this paper, the model is applied to study the fault diagnosis of reciprocating compressors and the result shows the validity of the diagnosis and realizes the potential fault diagnosis.
