Oil Well Fault Diagnosis Method Based On Electrical Parameter Time Series Analysis

Rod pumps are widely used in oil field exploitation projects,due to the structure of the equipment and its working environment,resulting in frequent failures,which makes the economic efficiency of oil low.Therefore,fault diagnosis for oil wells can understand the operation status of rod pumps.In the detection of operating conditions,it is currently more common to use the dynamogram based on oil wells,but the timeliness of this method is poor,resulting in the oilfield being unable to maintain a stable state for a long time.Moreover,due to the complex mechanism of oil wells,some fault types of oil wells are difficult to distinguish,that is,it is difficult to distinguish these faults through the dynamogram in a single cycle,which makes the fault diagnosis of oil wells more difficult.In order to solve the above problems,this paper proposes well fault diagnosis based on electrical parameter timing analysis.In this method analysis,the periodicity and trend of the well electric power of the well can be excavated,including the information of the well working conditions,and the well electrical parameters have good real-time,safety and stability.Therefore,in view of some problems existing in the dynamogram,a time-series fault diagnosis method based on oil well electrical parameters is proposed and its feasibility is verified.Firstly,the collected dynamogram and electrical parameters of the oil well are analyzed,and the causes of some working conditions of the well are analyzed in combination with the dynamogram.The traditional dynamogram-based fault diagnosis is abandoned,and the electrical parameters of the oil well are selected as the data,and the fault diagnosis of the oil well is carried out through data-driven fault diagnosis,combined with deep learning and electrical parameters.Before transferring the data to the model,it is necessary to preprocess the well electrical parameters,such as data cleaning and data transformation,to eliminate the useless data,and convert the preprocessed well electrical parameter data into GAF images and send them to the model.Secondly,fault diagnosis is carried out for the timing characteristics of the well.This paper introduces the basics of time series data,explains the oil well operating condition information contained in the electrical parameters,and understands the characteristics of the time series data and the time series characteristics contained in the electrical parameters,so that the oil well electrical parameter data can be applied to fault diagnosis.Through analysis,the DRSN model is selected for time series fault diagnosis,and the superiority of the model is proved by comparing with network models such as RNN,CNN and ResNet.Then,aiming at the timing problem of oil well electrical parameters,a fault diagnosis method based on deep residual shrinkage network(DRSN)model is proposed.The data is divided into a training set and a test set,the training set is used to train the model,and the test set is used to test the accuracy of the model,and at the same time select the parameters of the model,select the best model inside,and save it to the local for subsequent operations.Finally,by calling the network model described earlier,the fault diagnosis of field data is verified,and the feasibility of the proposed method is verified.In the experimental results,the timing analysis method based on oil well electrical parameters has good performance and great prospects in the fault diagnosis of oil wells.