Research on Electrical Parameter Fault Diagnosis Method of Oil Well Based on TSC-DCGAN Deep Learning

Due to the wide distribution and large number of oil Wells, it is extremely difficult to obtain oil well data, and the lack of electrical parameter data samples greatly restricts the development of intelligent diagnosis technology of oil well working conditions based on electrical parameters. In order to solve this problem, this paper proposes a fault diagnosis method based on time series transformation generative adversation network (TSC-DCGAN). Firstly, on the basis of analyzing the normal and typical fault working mechanism of rod pump, the simulation model of the whole process of mechanical production system is established, and the electrical parameter data for oil well fault diagnosis is generated, valuable data sets of real oil well electrical parameters are obtained. Secondly, in view of the limited data category generated by the mechanism model, the method of TSC-DCGAN is used to expand the electrical parameter data. Finally, a variety of deep learning network models are used to classificy the extended electrical parameter data. The EfficientNet deep learning network achieved the best classification effect, with an accuracy of more than 0.95. Experimental results show that the proposed method is effective, and we hope this method can promote the development of oil well intelligent diagnosis technology.