基于小波变换的桩基缺陷识别方法

Method for Cognizing Deficient Pile Based on Discrete Symmetry Wavelet Transform

在反射波法的桩基信号检测中主要归纳为检测幅值突变,当反射信号较弱时,用一般的时域分析法或频域分析法,就无法检测桩身质量问题.用小波分析就可以解决这一问题,小波分析具有时频局部化特点.从理论上说,桩基信号混有反映桩基缺陷的脉冲信号,通过对称小波变换后,在一定的尺度范围内其极值点对应原信号的突变位置或时刻.工程中信号分析如用连续小波变换产生的数据量多,且计算量大,分析复杂.而用离散小波变换可以有效地减少数据量,并能识别工程信号的奇异性(缺陷).论文通过一个在确定位置有微弱幅值变化,进行小波变换后(其中小波为二次B样条小波),可以清楚地识别幅值发生突变的时刻.同时也对工程实际中的一具体桩基用低应变动态检测的数据进行了分析,根据桩的破坏特征及离散小波变换后计算桩身发生缺陷的位置,可以确定本例中的桩有4处发生了离析.并且可以有效地对桩基缺陷进行识别.

Amplitude change is the main characteristic for pile signal detection by Reflection Wave Method （RWM）. If the reflected signal is weak, the traditional data processing methods are difficult to identify the deficiency of pile, such as time - domain analysis or frequency - domain analysis. Whereas wavelet analysis which has local time - frequency characteristics can deal with weak signal perfectly ; signals mixed with impulse signal which is regarded as deficiency in pile by Wavelet Transform （WT）,here,wavelet is secondary B spline wavelet. And the procedure is proposed to detect deficiency in pile. Two examples are illustrated in this paper：one numerical example has an explicit mathematic expression with a minute change in amplitude at certain position, and the minute change is detected by the proposed method;the other example is that the singularity of a practical engineering signal of a concrete pile is identified, using the tested low - strain dynamic data in pile to estimate the time of deficiency appearance by discrete wavelet analysis, then evaluating the deficiency position with the assumption invariable wave velocity in pile. Considering characteristics of various piles, the example pile has four isolations. So the proposed method is valid for cognizing deficient pile.