结构试验中一种高保真的数字滤波算法  

A High Fidelity Digital Filter Arithmetic in Structural Experiment

作　　者：周英武[1] 王苏岩[1] 李宏男[1] 

机构地区：[1]大连理工大学土木水利学院,大连辽宁116024

出　　处：《重庆建筑大学学报》2006年第6期42-46,共5页Journal of Chongqing Jianzhu University

基　　金：国家自然科学基金项目(50378711)

摘　　要：结构试验中在采集诸如荷载挠度曲线等数据时,由于受到外界环境不可预测的信号的干扰,往往会使采集的数据产生很大的漂移,而且漂移是随机的,不可预测的,这就严重歪曲了试验的真实曲线,给试验的分析处理带来了困难。本文根据随机干扰信号的特点,采用了一种三阶段的数字滤波算法,每个阶段采用不同的滤波思想,逐步地将干扰信号过滤出去,还原结构试验曲线的真实面貌,并通过两个具体不同的实例,验证了算法的可行性。During data collection, the experimental data such as curve of force versus displacement are usually disturbed by the environmental uncertain signals, which cause the data to drift greatly. Since the drift is stochastic, uncertain, and has distorted the real experimental curve severely, it brings lots of difficulties in analyzing the experiment results. In this paper,based on the characteristic of the stochastic disturbance signals, a three - phase digital filter arithmetic is given,in which the disturbance signals are filtered in each phase gradually, according to the different filter methods, and then the real curve is reappeared again. Two idiographic and different instances are adopted to prove the validity of the arithmetic at the end.

关 键 词：数字滤波 随机干扰 改进的一阶差分 中位数 滑动平滑 

分 类 号：TU317[建筑科学—结构工程]