贝叶斯估值理论在小子样批次设备检测中的应用

The Application of Bayesian Estimate Theory in the Extreme-small-sample Tests of Batch Equipments

本文针对机载无线电设备故障诊断定位率低的问题,提出利用先验故障信息指导批次定检的思想.通过研究批次定检设备的失效规律,运用贝叶斯理论估计小子样定检设备各模块的故障概率,并且提出了故障检测的4种指导方法.仿真试验结果表明可显著提高故障的一次定位准确率,减少故障查找时间,提高故障诊断效率.

To resolve the problem of low locating rate in fault diagnosis of air-borne radio equipment, the idea was proposed to using preceding fault information in the tests of batch equipments. By studying the failure rules of batch equipments, the fault probability under extreme-small-sample conditions was estimated using Bayesian theory. Four methods were adopted to direct fault detecting. The simulation results showed that the strategy of this paper could improve the rate of locating faults correctly in one-time detecting progress and reduce the fault searching time, and can enhance the efficiency of fault diagnosis.