摘要
本文简要介绍了剔除坏点数据的几种经典方法,并对动态测试中数字时序信号的坏点剔除方法进行了探讨。根据动态测试中数字时序信号的特点提出了一种剔除坏点的新方法———绝对均值法,并给出了绝对均值法中阀值W确定的范围。作者采用绝对均值法和莱因达(PaйTa)法分别对三组实测数据进行坏点数据的剔除处理,对处理结果的分析对比表明运用绝对均值法剔除动态测试数据中的坏点数据比莱因达法效果更好。
This paper discussed a few traditional methods to remove singular points in time series signals.According to the characters of the time series signals by dynamical measure, a new method for removing the singular points, absolute mean value method, was put forward, in which the range of the threshold W was determined. Compared the absolute mean value method with PaЙTa method in processingthree groups of datum by measure, it is found that the absolute mean value method works more accurately and effectively.
出处
《中国测试技术》
2006年第1期47-49,82,共4页
CHINA MEASUREMENT & TESTING TECHNOLOGY
关键词
坏点剔除
绝对均值法
数字时序信号
动态测试
数据处理
Removing singular points
Absolute, mean value method
Time series signal
Dynamic testing
Datum processing
