Fault Diagnosis Research of Submarine Casing Cutting Robot for Abandoned Oil Wellhead

Abstract

The effectiveness of a submarine casing cutting robot is mainly influenced not only by its operational but also by its reliability and safety. In this paper, fault diagnosis research of this cutting robot is evaluated using the Bayesian network. A methodology of transforming the fault tree model into Bayesian network model is used. The fault tree model is established simply and conveniently. Bayesian network can address interesting questions allowing both forward and backward analysis. Combining the merits of two methods, the causes of failures, the occurrence probabilities and the importance of various components are analyzed based on the Netica software. The results show that the robot has high reliability and should be paid attentions to the research of feeding mechanism and the discharge gap detection circuits.