摘要
运用因变量估计自变量及其可能取值范围的逆回归分析方法代替传统的回归分析方法,推导出回弹值和声速值的逆回归模型数据融合测强曲线。运用区间估计方法代替传统的点估计方法,导出结构混凝土强度的置信上限、下限和置信区间。给出混凝土强度的判定标准,在给定高置信度下对结构混凝土强度做出合格、不合格和待定结论的定量判断。该测强曲线比任何单一方法的测强曲线的精度高,减小偶然误差,能够消除无损检测中无法通过仪器校准的系统误差,并且解决了回弹值和声速值不同量纲的数据融合问题。
By using the inverse regression method, which uses dependent value to estimate independent value and its span instead of the traditional regression method, the inverse regression model data fusion strength curve of rebound and velocity value is deduced. Meanwhile, by using interval estimation instead of the conventional point estimation, the lower and upper confidence limits, the confidence intervals of structure concrete strength are derived. Moreover, the judgment standard of concrete strength is given. At the given high degree of confidence, the quantitative judgment of structure concrete strength is made, that is acceptance, rejection or undetermined. This strength curve is more accurate than any other curve got by using single method. It can reduce accidental error and eliminate system error that canr t be calibrated by instrument in nondestructive test. Besides, the data fusion problem of different dimensions of rebound and velocity value is resolved.
出处
《中国铁道科学》
EI
CAS
CSCD
北大核心
2006年第1期43-47,共5页
China Railway Science
基金
河北省唐山市科技局项目(04363901B-8)
关键词
结构检测
无损检测
超声回弹综合法
数据融合
回归分析
逆回归分析
Structure test
Nondestructive test
Ultrasonic-rebound combined method
Data fusion
Regression analysis
Inverse regression analysis