Study On The Real-time Fault Diagnosis And Analysis System Of Rod Pumped Well

The fault of oil well is an important factor that affects the oil field production in the way of sucker rod pump production,and processing well faults timely can effectively improve the production efficiency of the oil field.Recently there are two ways to judge the well fault,one is monitoring of production data changes online,another one is discrimination of the faults based on indicator diagram,the former cannot monitor well faults in the round,and the latter cannot be determined well failure in time.Therefore,combined with the theoretical knowledge of oil production engineering technology,computer visual simulation and fault diagnosis,the fault diagnosis system has been further studied,and the real-time fault diagnosis and analysis system of the rod pumped well has been developed.Carrying out the functions framework of the software by software requirement analysis,which mainly includes two modules wells online fault diagnosis and well analysis.Through Visual Studio as a tool for software development,using Access database for data management,the use of OpenGL technology to achieve software visualization simulation function,and finally completed programming of the fault diagnosis software system.In the online fault diagnosis module,which preprocesses indicator diagram data,shows indicator diagram curve by Freeman chain code,then extracted the regional characteristics of indicator diagram according to regional similarity.The process does not need to solve the wave equation about pump indicator diagram,for which improving the speed of calculation.The BP neural network is constructed,which is applied in the method of regional similarity classification,combined with the traditional area method,which forms a comprehensive fault diagnosis method.The mathematical model and wave equation of sucker rod system are established in the oil well optimization analysis module,and the conversion of the hanging point indicator diagram to pump indicator diagram is completed.The Freeman chain code of the typical fault indicator diagram is summarized and analyzed,and get the fault gene chain code of the oil well fault indicator diagram,simulated the oil well fault condition indicator diagram based on that,which made the calculation of the oil well liquid production,the stress intensity of the check rod string,the optimization of the rod string combination design,the calculation of pump efficiency composition,calculation of the oil well friction and the quantitative and qualitative analysis of the fault degree and the others functions come true.By using the grey system forecasting model,the forecast and analysis of the data of future indicator diagram is completed.To realize the visualization of dynamic parameters of the oil well,the visualization of indicator diagram contrast display,the visualization of neural network diagnosis results and the 3D visualization of the multi indicator diagram evolution.