Dataset of Intrusion Events in Oil and Gas Pipelines

Published: 29 July 2025| Version 1 | DOI: 10.17632/w7nzxs593c.1
Contributor:
Han Qin

Description

This dataset is divided into test data, training data, and dedicated label data sections. The oil and gas pipeline intrusion events are classified into three categories: manual tapping (labeled as 0), mechanical excavation (labeled as 1), and human trampling (labeled as 2). The test data and training data are used for model evaluation and training, respectively. During the training process, labels can also be manually removed for unsupervised learning.

Files

Institutions

  • Xihua University

Categories

Engineering Database, Optical Fiber Sensing, Deep Learning

Licence