为满足实际工程、生产应用中专业及非专业人员对故障光纤激光器快速修理、修复的需要, 研究了一种能够实现快速故障模式及故障原因分析系统。首先, 采集实际使用与实验中损坏激光器的故障模式与故障原因记录为数据样本。接着, 利用故障树与故障模式影响及危害性分析模型建立光纤激光器扩展故障树, 即构建起故障模式与故障原因的对应关系。然后, 利用贝叶斯网络分析并得出各故障模式与故障原因的先验概率与后验概率计算方法, 再使用Python语言编写交互式窗口, 并完成对数据库数据的调用与相应概率计算。最后, 通过交互式界面输出当前故障对应的可能故障原因及发生概率, 从而实现对光纤激光器故障的快速分析。程序模拟耗时小于1 s, 模拟结果与统计结果相吻合, 基本满足指导相关人员对故障快速排查并修理的要求。

In order to meet the needs of professional and nonprofessional personnel about fast repairing the fault fiber laser in the practical engineering and production applications, a fast fault mode and fault cause analysis system was reaearched. First, the fault mode and fault cause of the damaged laser in actual use were collected as samples. The extended fault tree of fiber laser was established by the fault tree analysis and failure mode effect criticality analysis, i.e. the corresponding relationship between failure mode and fault cause was constructed. Then, Bayesian network was used to analyze and obtain the prior probability and posterior probability of each failure mode and fault cause. Using Python to write interactive window and call database data, the corresponding probability was calculated. Finally, the possible fault mode and correspond probability were output through interactive interface, so as to achieve rapid analysis of fiber laser fault. The analysis system simulation results coincide with the reality, calculating less than 1 s, so the system can basically meet the relevant personnel to guide the rapid troubleshooting and repair requirements.