分布式光纤桩基检测里特征点压缩算法的应用研究

Research on the application of Feature Point Compression Algorithm in Distributed Fiber Optic Pile Inspection

在开展桩基分布式检测工作时,常常用布里渊散射光时域反射技术(BOTDR)测量桩基,对数据的平滑去噪处理,是此工作中重要的一步。以某桩基为例子,借助特征点压缩算法,对检测数据进行平滑去噪处理,分析这种数据处理方法的特点和在实际中的应用效果。研究结果表明:数据的去噪会受到压缩算法的特征点阈值影响,参数的选择应结合实际情况和具体数据来确定;以不同的指标对平滑处理效果进行分析时,得出的压缩算法最优次数也不同;从摩阻力和轴力分布图来看,实际应用效果和桩体荷载传递规律以及国家相关规范基本一致,可以将荷载作用下桩体的实际受力情况准确地反映出来,效果良好。

The Brillouin scattered light time domain reflection technique(BOTDR)is often used to measure pile foundations in distributed inspection work,and smoothing and denoising of the data is an important step in this work.In this paper,we take a pile foundation as an example,and analyze the characteristics of this data processing method and its application in practice by smoothing and denoising the inspection data with the help of feature point compression algorithm.The results show that:the denoising of the data is affected by the threshold value of the feature points of the compression algorithm,and the selection of parameters should be determined with the actual situation and specific data;the optimal number of compression algorithm is different when the smoothing effect is analyzed with different indexes;from the frictional resistance and axial force distribution diagrams,the actual application effect is basically consistent with the pile load transfer law and relevant national codes,and the actual force of the pile under load can be The actual pile force under the action of the load can be accurately reflected,and the effect is good.