深水防喷器液控系统故障诊断研究

Research for Fault Diagnosis of Deepwater Blowout Preventer

采用贝叶斯方法对深水防喷器液控系统建立典型诊断网络模块,依据专家知识以及历史统计数据确定贝叶斯网络中的节点参数,包括节点的先验概率和节点之间的条件概率。对防喷器关闭诊断贝叶斯网络进行了试验,诊断结果基本符合实际。证明了应用贝叶斯网络方法对深水防喷器液控系统进行故障诊断研究的可行性。

The typical diagnosis network module for deepwater BOP hydraulic control system that base on Bayesian methods is established.On the basis of expert knowledge and historical statistics to determine the Bayesian network node parameters,including the nodes’priori probability and conditional probability between nodes.The diagnostic Bayesian network of BOP close is tested, and the results are more accurate diagnosis,in line with reality.Demonstrate the feasibility of the application of Bayesian networks for deepwater BOP hydraulic control system fault diagnosis.