摘要
在航天器部组件产品试验时存在着海量闲置数据,但由于缺少有效的数据挖掘手段,型号产品的寿命和质量一致性一直得不到有效评估。本文给出了航天器部组件产品实时寿命预测方法和质量一致性评价方法的实施流程,并分析了各类退化预测建模方法和包络建模方法的适用条件。案例验证表明,实时寿命预测可以采用产品性能的实时监测信息,建立产品退化特征预测模型,实现对加速寿命的动态预测;质量一致性评价可以采用历史产品试验信息,构建成功包络线,实现对被试产品是否满足质量一致性要求的有效判别。
There are large volumes of data at rest in spacecraft component tests.However,due to the lack of effective data mining methods,the lifetime of products and quality consistence cannot able to be evaluated effectively.This paper proposes a real-time prediction flow of the accelerated life and an evaluation flow of quality consistence.Moreover,various modeling methods of degradation prediction and envelopment analysis are analyzed.Two case studies are carried out,and the results show that the real-time prediction method can establish a degradation model for spacecraft components by using the real-time performance monitoring information and realize the dynamic prediction of accelerated lifetime;the quality consistence evaluation method can build the successful envelope with the test information of historical products,and effectively determine whether the quality consistency of the test products meets the requirements.
作者
秦泰春
刘守文
周月阁
庞博
Qin Taichun;Liu Shouwen;Zhou Yuege;Pang Bo(Beijing Key Laboratory of Environment&Reliability Test Technology for Aerospace Mechanical&Electrical,Beijing Institute of Spacecraft Environment Engineering,Beijing,100094,China)
出处
《南京航空航天大学学报》
EI
CAS
CSCD
北大核心
2019年第S01期149-154,共6页
Journal of Nanjing University of Aeronautics & Astronautics
关键词
大数据
试验
数据挖掘
动态寿命预测
质量一致性
big data
laboratory test
data mining
life prediction
quality consistency