性能退化与寿命数据融合的贝叶斯方法研究

Research on Bayes Parameter Evaluation Based on Performance Degradation and Life Data Fusion

针对高可靠性航空产品寿命失效数据样本量小,传统评估方法可信度低的难题,提出基于性能退化数据与寿命失效数据融合的可靠性评估方法。然后用失效机理分析、建模方法选择、失效阀值确定、置信区间求解和数据融合5个关键步骤具体阐述了该方法。最后以飞机液压泵为例,使用该方法进行性能退化数据与寿命数据的融合计算。分析表明,该方法充分利用了性能数据和寿命数据,比传统的方法更加合理,评估结果更加精确。

Aiming at the problem that the life failure data of high reliability aviation product is short, and the reliability of traditional evaluation way is low, the data fusion algorithm based on performance degradation data and life failure data is put forward to evaluate the reliability. Then the way is described by five critical steps, namely analysis of failure mechanism, selection for ways building models, the failure threshold selection, confidence interval solving, and the data fusion. In the end, the hydraulic pump on airplane is taken as an example to use this method to calculate the performance degradation data and life data. The results taking full advantage of performance data and life data show that this way is more reasonable and accurate than traditional ways.