小波变换与近似熵原理在桩基检测中的应用

Application of Wavelet Transform and Approximate Entropy in Pile Foundation Inspection

桩基混凝土质量评级常采用PSD多参数综合方法,但由于随机扰动的存在和混凝土本身的不连续特性,在检测过程中会混入奇异信号,导致对缺陷产生漏判或者误判。文章利用小波变换对桩基的疑似缺陷信号进行分解,同时引入近似熵定量表征信号的无序程度。结果表明缺陷信号经过小波分解后,在高频段上信号幅值衰减且波形杂乱、近似熵值大。该方法特别适合判别疑似缺陷,工程实践证明了其在桩基检测中的有效性。

PSD multi parameter approach is often used in concrete pile quality rating. However, in testmg process, because of random disturbance to the inspection signals and discrete features in the concrete pile, singular signals may be produced and lead to miss or misinterpretations of defects. In this paper, suspected defect signal was decomposed by wavelet transform, and the disorder of signal was quantitatively described by introduction of approximate entropy. The results showed that under the wavelet decomposition, the amplitude of defect signal was attenuated in the high frettuency hand with a clutter waveform and larger approximate entropy. The present method was be fitted for confirming suspected defect and showed good effectiveness in the practical pile foundation inspection.