摘要
随着空间科学技术的不断发展,人们进行的空间活动日益增多。与此同时,空间碎片的数量也随之急剧增多,这使得在轨航天器遭受空间碎片等动能颗粒撞击可能性不断增加,航天器发生密封结构泄漏的可能性越来越大。航天器泄漏不仅会造成航天活动的失败,更会威胁航天员的生命安全,因此,在轨航天器真空泄漏声发射在线检测技术研究具有十分重要的意义。本文以声发射技术为基础,利用精细平均功率谱分析法对不同泄漏孔径下产生的声发射信号特征进行提取分析,并与改进的混沌-BP神经网络模式识方法相结合实现对信号进行分类,从而实现泄漏检测以及漏孔大小评估。通过真空泄漏声发射检测实验验证,本方法最小可检0.4 mm直径的漏孔,满足在轨航天器微弱泄漏检测的需求。
Here,we addressedthe online detection of vacuum leakageby acoustic emission,resulted from the accidental impact of space debris and/or high energy particleson in-orbit spacecraft. In the ground simulation,first,the characteristic parameters of the acoustic signals,emitted from different leakage apertures,were sampled and evaluated in average power spectrum analysis; next,the acoustic emission signals involved were classified in the modified method of chaotic back propagation neural network model; and finally,the influence of the aperture size of leakage hole and characteristics of the sampling distance on the leakage detection precision was investigated in simulation and with the lab-built test platform. The simulated and measured results were found to be in good agreement. We concluded that the novel technique is capable of accurately detecting the smallest leakage aperture down to 0. 4mm,meeting the requirements for weak leakage online detection of on-orbit spacecraft.
出处
《真空科学与技术学报》
CSCD
北大核心
2017年第6期649-653,共5页
Chinese Journal of Vacuum Science and Technology
关键词
泄漏检测
声发射
特征参数
混沌算法
BP神经网络
Leak detection
Acoustic emission
Characteristic parameters
Chaos algorithm
BP neural network
