多尺度多传感器约束的声发射源定位研究  

Acoustic emission positioning with multi-scale and multi-sensor constraints

作　　者：王平[1] 唐力伟[1] 张晓涛[1] 王晓明 

机构地区：[1]军械工程学院火炮工程系 [2]63981部队

出　　处：《中国工程机械学报》2015年第5期461-465,共5页Chinese Journal of Construction Machinery

摘　　要：针对突发型声发射信号直接互相关所求时延不准确,导致源定位结果可靠度较低的问题,提出一种结合互累积量时延估计与多传感器相互约束和信号多尺度分解的声发射源定位方法.利用多传感器相互约束的时延矢量封闭准则,在信号多尺度分析的子频率带内筛选最精准的时延,获取高可靠度的源定位结果,通过声线定位与平面定位试验对该方法进行验证,结果表明:互累积量方法所求时延准确,线定位与面定位结果的最大相对误差分别为1.02%和11.7%.Owing to the inaccurate time delay from direct signal correlations of emergent acoustic emissions,the low reliability of source positioning exists.In this study,an integration of crossaccumulation time delay estimation,multi-sensor constraint and multi-scale decomposition is proposed for source positioning.Based on the time delay vector close criterion of multi-sensor constraint,the most accurate time delay is selected from sub-frequency brands via multi-scale analysis to obtain teh high reliability result.By applying acoustic line and plane positioning methods for verification,it is detected that the cross-accumulation method possesses higher accuracy,while the maximum relative errors from line and plane positioning methods are relatively 1.02% and 11.7%.

关 键 词：声发射 时间延迟 互累积量 多尺度源定位 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]