基于多信号温度模型的机载电子板卡故障检测研究

Research of Fault Diagnosis System of Airborne Circuit Board Based on Multi-signal Temperature Model

对机载电子板卡的红外故障检测进行了深入研究,通过将机载电子板卡正常工作时的红外温度图像与故障时的进行对比,建立了系统的多信号温度模型,针对传统多信号模型的缺点,提出一种基于贝叶斯网络的建模方法,并给出贝叶斯网络模型的测试性分析方法与指标计算方法,得到模型相关矩阵并进一步进行故障诊断分析;结果证明该模型考虑了故障与测试之间的不确定因素,其计算结果更符合真实情况,为检测电子元器件故障及老化程度提供了一种有效途径。

The infrared fault detection of airborne electronic board is studied in detail. Through the comparison between the infrared temperature image of the board work properly and the failure, established a system of multi--signal temperature model, against the disadvantages of traditional multi--signal temperature model, proposes a testability modeling method based on Bayesian network, and gives the test model analysis method and the index calculation method of the Bayesian network model, obtain the correlation matrix model and the analysis of the fault diagnosis. Results show that the model considering the uncertain factors between the fault and test, the calculated result conforms to the real situation, provides an effective way to detect electronic components failure and aging degree.