钢轨表面及亚表面检测数据集

钢轨表面、亚表面伤损会影响行车安全，使用漏磁技术进行检测。本数据集来自深圳地铁世界之窗–侨香区间、后海-侨城北区间，设备采用与大型钢轨探伤车GTC-80相似的漏磁检测设备，使用滑靴式探头，分为顶面滑靴和侧面滑靴，并合理规划滑靴内部的传感器数量和位置，使检测范围覆盖整个钢轨轨头顶面、主要包括踏面传感器、轨距角传感器、磁化器、信号调理放大电路、供电电源、工控机和采集卡，对传感器获取的电压信号进行处理和采集。检测过程为检测探头部分的磁芯施加直流激励源，磁化被测钢轨试件，霍尔传感器阵列检测被测钢轨试件的漏磁场变化情况，获取漏磁信号。经过调理电路放大、低通滤波后，计算机的采集卡把采集到的漏磁信号处理分析，进行漏磁成像显示，反映出被测钢轨试件表面缺陷分布和特征。数据集包括各个通道检测到的漏磁场原始数据，为自定义格式，数据量约50G。

Surface and subsurface defects on steel rails pose threats to traffic safety, which are detected using magnetic flux leakage (MFL) technology. This dataset is collected from the Window of the World-Qiaoxiang interval and Houhai-Qiaocheng North interval of Shenzhen Metro. The adopted equipment is a magnetic flux leakage detector similar to the large-scale rail flaw detection vehicle GTC-80, which uses shoe-type probes divided into top shoe and side shoe. The quantity and position of sensors inside the shoes are reasonably planned to ensure the detection range covers the entire top surface of the rail. The system mainly includes tread sensors, gauge corner sensors, magnetizers, signal conditioning and amplifying circuits, power supplies, industrial personal computers (IPC), and acquisition cards, which process and collect the voltage signals acquired by the sensors. The detection process is as follows: a DC excitation source is applied to the magnetic core of the detection probe to magnetize the tested rail specimen, and the Hall sensor array detects the changes in the leakage magnetic field of the tested rail specimen to obtain MFL signals. After being amplified by the conditioning circuit and low-pass filtered, the collected MFL signals are processed and analyzed by the computer's acquisition card, and then displayed through MFL imaging, which reflects the distribution and characteristics of surface defects on the tested rail specimen. The dataset contains raw leakage magnetic field data detected by each channel, which is in a custom format, with a total data volume of approximately 50 GB.

搜集汇总

背景与挑战

背景概述

该数据集是一个钢轨表面及亚表面缺陷检测数据集，采用漏磁检测技术，通过滑靴式探头和霍尔传感器阵列采集深圳地铁特定区间的钢轨漏磁信号，数据量约为12.48GB，包含原始检测数据，主要用于钢轨伤损的安全评估和研究。