摘要
提出了一种数字滤波的新算法。为了度量测量数据的可靠性程度,该算法引入了测量数据的支持量和有效支持量等概念。确定支持量的关键在于滤波阈值的选取,算法对测量数据进行分类,并根据不同的分类数给出不同的滤波阈值,所给出的滤波阈值能在不损失有用信息的情况下,抑制异常数据影响的扩散。而有效支持量的引入,可以使异常数据的残余影响进一步被消除。数据实验的结果证明：与同类滤波算法相比,当连续出现一个或多个异常数据及测量真值出现跃变时,该算法具有更强的抗干扰和快速处理能力。
A new digital filter algorithm is proposed. The concept of support and effective support measure is introduced to survey the degree of measurement data reliability. How to choose the filter threshold is a key to determining the support measure. First, measurement data is classified ,and then different filter thresholds are determine according to the different classification numbers. It can repress the proliferation of abnormal data influence on the case of not losing useful information. Because of the introduction of the effective support measure,it can further remove the remnant influence of abnormal data. The data experiment confirms that when some abnormal data appear continuously and the true value jumps,it has good anti-interference and fast tracking ability as compared with the same kinds of algorithms.
出处
《数据采集与处理》
CSCD
北大核心
2006年第2期234-238,共5页
Journal of Data Acquisition and Processing
基金
浙江省光纤通信技术重点研究实验室资助项目
关键词
异常数据
数字滤波
阈值
聚类分析
abnormal data
digital filter
threshold
clustering analysis
