8 / 2021-08-14 10:59:05

A dynamic Bayesian network-based intermittent fault diagnosis methodology for downhole motor

Dynamic Bayesian network (DBN); Intermittent faults (IFs); Fault diagnosis; Downhole motor (DM)

As the results of fault diagnosis vary with the working conditions and the performance degradation of downhole motor, an intermittent fault diagnosis methodology based on a dynamic Bayesian network (DBN) is proposed in this study. The methodology simplifies the working conditions of the downhole motor to establish Markov models. The Markov models that simulate the transition of fault types are combined with the DBN models that simulate the dynamic degradation process of components for fault diagnosis. The proposed methodology is used to identify the malfunctioned components and discriminate the types of faults, e.g., intermittent faults and permanent faults. Three application cases are used to confirm the effectiveness of the methodology for intermittent faults in degradation components of the downhole motor.