管道威胁事件数据集

本数据集主要面向油气管道安全监测领域的智能威胁识别算法研究与验证需求建设。在真实的油气管网运营中，第三方破坏等威胁事件数据具有偶发性、获取成本高、危险性大等特点，导致可用于算法训练与评估的有效样本极为稀缺。为解决这一行业痛点，本数据集应运而生。该数据集基于国家重点研发计划“跨地域复杂油气管网安全高效运行状态监测传感系统及应用”项目课题成果“油气管道传感器及系统综合测试平台”产生，依托浙江衢州试验场，利用分布式光纤传感系统进行全面采集。数据的产生方法兼具标准化与高保真两大特性：一方面，通过自研的标准化振动装置，生成特定频率、可重复的基准振动信号；另一方面，动用真实的工程机械（挖掘机）与人工作业（铁锹挖掘），高保真地复现了现实世界中的威胁场景。所有数据均由分布式光纤传感系统采集，数据采样频率为1500Hz，空间分辨率为400米，数据量4GB。 数据集总体量包含5个独立的二进制（.bin）数据文件，其中数据文件1和2为标准化振动装置产生的信号，文件3为人工挖掘信号，文件4和5为挖掘机作业期间的综合信号。数据格式为单精度浮点数，采用大端字节序，并包含120字节的元信息头部。本数据集的发布，旨在为管道安全领域的科研人员提供一个高质量、标注清晰、工况丰富的基准测试库，有力支撑新型威胁事件检测、分类与识别算法的开发、训练与性能评估工作。

This dataset is developed to meet the research and verification needs of intelligent threat recognition algorithms for oil and gas pipeline safety monitoring. In actual oil and gas pipeline network operations, data of threat events such as third-party damage are sporadic, high-cost to acquire, and highly risky, leading to an extreme shortage of valid samples available for algorithm training and evaluation. To solve this industry pain point, this dataset was constructed. This dataset is generated based on the results of the "Comprehensive Test Platform for Oil and Gas Pipeline Sensors and Systems", a sub-project of the National Key R&D Program of China's "Safe and Efficient Operation Status Monitoring Sensing System and Application for Cross-regional Complex Oil and Gas Pipeline Networks", and was comprehensively collected using a distributed optical fiber sensing system at the Quzhou, Zhejiang test site. The data generation method has two key characteristics: standardization and high fidelity. On one hand, a self-developed standardized vibration device is used to generate benchmark vibration signals with specific frequencies and repeatability; on the other hand, real engineering machinery (excavators) and manual operations (shovel digging) are employed to high-fidelity reproduce real-world threat scenarios. All data is collected by the distributed optical fiber sensing system, with a sampling frequency of 1500 Hz, a spatial resolution of 400 meters, and a total data size of 4 GB. The dataset includes 5 independent binary (.bin) data files, where files 1 and 2 contain signals generated by the standardized vibration device, file 3 contains manual digging signals, and files 4 and 5 contain comprehensive signals collected during excavator operations. The data format is single-precision floating-point numbers with big-endian byte order, and includes a 120-byte metadata header. The release of this dataset aims to provide researchers in the field of pipeline safety with a high-quality, well-annotated, and diverse operational conditions benchmark test library, which strongly supports the development, training and performance evaluation of novel threat event detection, classification and recognition algorithms.

搜集汇总

背景与挑战

背景概述

该数据集是一个面向油气管道安全监测的威胁事件模拟测试数据集，主要用于智能威胁识别算法的研究与验证。数据集基于国家重点研发计划项目，在浙江衢州试验场利用分布式光纤传感系统采集，包含标准化振动、人工挖掘和挖掘机作业等多种高保真威胁场景数据，数据采样频率为1500Hz，空间分辨率400米，总数据量4GB，为算法开发提供高质量基准测试库。