基于贝叶斯网络Noisy-Or模型的低压漏电保护器机械组件故障诊断应用

Fault diagnosis for mechanical components of leakage protection devices based on Bayesian networks Noisy-Or model

机械组件作为低压漏电保护器中最重要的部件之一,对整个产品的最终性能和质量起到了至关重要的作用。由于机械组件受到的影响因素较多,往往造成其故障率较高,合格率也经常处于不稳定的状态,因此,有必要研究机械组件模型以及故障诊断方法。论文提出了基于贝叶斯网络的低压漏电保护器机械组件故障诊断方法,通过对机械组件不良品进行拆解分析并获取生产线与实验室数据,完成贝叶斯网络Noisy-Or模型的构建,最终通过故障原因后验概率大小判断主要故障原因,为后续的故障改进方案提供参考。最后,结合案例对比分析得出使用贝叶斯网络Noisy-Or模型可大大减少所需条件概率的数量,并验证了模型的可行性与实用性。

As one of the most important parts of low-voltage leakage protection devices, mechanical component plays a crucial role in the final performance and the quality of the devices. The pass rate is always in a state of in- stability because it is influenced by too many elements, so it has significant application value for improving the pro- duction efficiency of the enterprise by building a new model for the fault diagnosis of mechanical components. In this paper, we propose a model for the fault diagnosis of mechanical component of leakage protection devices based on Bayesian networks. The Bayesian networks Noisy-Or model is built by tearing down and analyzing the mechani- cal component and getting the data of laboratory. Finally, the main reason of fault is judged by posterior probability and a reference for the improvement program is provided. A case is given to demonstrate that the Bayesian networks Noisy-Or model can greatly reduce the number of conditional probability, and the effectiveness and practicability of the method are verified.