Research data supporting chapter 'Unsupervised Representation Learning for Monitoring Rail Infrastructures with High-Frequency Moving Vibration Sensors' of Dissertation 'AI Solutions for Maintenance Decision Support in Railway Infrastructure'

The data and codes were prepared and uploaded to 4TU.ResearchData by Wassamon Phusakulkajorn to support the results in Chapter 4 (Unsupervised Representation Learning for Monitoring Rail Infrastructures with High-Frequency Moving Vibration Sensors) of her dissertation. This chapter has been submitted for publication as Phusakulkajorn, W., Zeng, Y., Li, Z., Núñez, A., Unsupervised Representation Learning for Monitoring Rail Infrastructures with High-Frequency Moving Vibration Sensors.  In this research, we develop an unsupervised representation learning methodology to automatically capture the dynamic responses of rail infrastructures and provide insights into the underlying characteristics of their conditions. The objective is to address the challenge when high-frequency vibration signals are obtained in new environments where prior knowledge or reference information about infrastructure conditions is unavailable or very limited. The data used contain validated axle box acceleration data for rail defect detection and train-borne laser Doppler vibrometer data for rail fastener monitoring. The implementations are done in Python jupyter notebooks and MATLAB in which (.ipynb, .py) and (.mat) files are analytical solutions and (.eps) and (.jpg) are figures used in the main manuscript.

Wassamon Phusakulkajorn整理并制备了本数据集与代码，并上传至4TU.ResearchData，用于支撑其博士论文第4章的研究结果——该章节标题为《无监督表征学习（Unsupervised Representation Learning）：基于高频移动振动传感器的铁路基础设施监测》。该章节已以Phusakulkajorn, W.、Zeng, Y.、Li, Z.、Núñez, A.为作者，提交至学术期刊发表。本研究提出一种无监督表征学习方法，可自动捕捉铁路基础设施的动态响应，并揭示其服役状态的内在特征。本研究旨在解决在基础设施状态先验知识或参考信息缺失、且极为有限的全新应用场景中采集高频振动信号时面临的挑战。本数据集包含经验证的轴箱加速度数据（用于铁路缺陷检测）以及车载激光多普勒测振仪（laser Doppler vibrometer）数据（用于铁路扣件监测）。本研究的实现基于Python Jupyter笔记本与MATLAB平台：其中.ipynb、.py格式文件为分析代码，.mat格式文件为分析结果数据，.eps与.jpg格式文件为主投稿手稿中使用的插图。