Fault Diagnosis of Indicator Diagram of Pumping Well Based on Stochastic Configuration Network

Abstract

In China’s oil exploitation, rod pumping wells occupy an important position. Once the pumping well breaks down, the oil production work will not be carried out in an orderly manner, which will affect the progress target and cause certain safety accidents in serious cases. Therefore, accurate fault diagnosis of pumping wells is a very necessary work. According to the coordinate points of oil well data acquisition, this paper carries out normalization processing, uses wavelet transform and singular value decomposition (SVD) to reduce noise, then draws the image, extracts the gray level co-occurrence matrix (GLCM)and contour features, and uses stochastic configuration network (SCN) to model the typical fault diagnosis of rod pumping wells. Finally, an example is used to verify the correctness of this method. Experiments show that the system has a high fault recognition rate, which verifies the efficiency of SCN classification. It can identify faults faster and more accurately in actual oilfield projects, and is of great significance to improve oil well production.