Design Of Fault Diagnosis Platform For Coal-bed Methane Gathering Process And Research On The Fault Diagnoses For Compressor

With the persistent development of national economy, China’s demand of oil and natural gas increases dramatically, and contradiction between energy supply and demand is more obvious. In recent years, as one of oil and gas energy, the coal bed methane(CBM) has been paid great attention to by the Chinese government. So the further development and utilization of domestic CBM resources is the most realistic and effective way to solve the contradiction between energy supply and demand. CBM exploration and gathering process is relatively complicated in the domestic, including:gas production at low-pressure in wellhead, centralized utilities delivering CBM in gas-gathering station, centralized processing in booster station. To keep all equipments in the normal operation is very important in CBM gathering process. When a failure happen, it may affect the production and transmission standards of CBM, even causes a chain reaction of the whole production process. So the whole process can not run normally and even the equipment will be damaged as a result of the failure.The research indicates that failure often has happened in wellhead and booster station in Coal bed methane (CBM) gathering process. To provide guarantees for the production and transmission standards of CBM, these faults must be monitored, diagnosed and handled by operating personnel in time. So to establish the fault diagnosis system in CBM gathering process for fault monitoring and diagnosis has a realistic significance in this paper.Based on technological process in coalbed methane (CBM) gathering process and fault diagnosis methods being applied and researched in domestic and international, this paper has designed a fault diagnosis system for CBM gathering process in Panhe gas extracation based on hybrid-programming with C#and MATLMB with Visual Stdio.NET2008, SQL Server2005and ADO.NET database access technology.This paper mainly studied fault diagnosis of compressor in the booster station, firstly we have analyzed the working mechanism and the failure mode of DF-5/10-40reciprocating compressor, and studied a variety of intelligent fault diagnosis methods. Based on complexity, uncertainty and potential fuzzy relation between faults and symptoms, fuzzy associative memory neural network model which combines the fuzzy theory and neural network theory, was constructed for fault diagnosis of compressor. This model implements intelligent fault diagnosis and self-learning of the knowledge database. At last, concrete examples has been adopted in experiment to test the mechanism of reasoning, significantly improve the accuracy and scalability of fault diagnosis system of the reciprocating compressor.