管廊管线故障声纹数据集

管廊管线故障声纹库数据在试点管廊（容西管廊豪丹段综合仓与讯电仓；容东管廊205段能源仓，水信仓与高压仓；容东管廊207段水信仓）采集与实验室管廊试验台中采集。在试点管廊中使用声纹传感器与录音主机采集管廊中的巡检脚步声纹与风机工作时的声纹；在北京交通大学唐山研究院载运工具运行安全实验室中通过实验平台采集管廊失效相关的声纹（管廊供电线路失效声纹；管廊供水管路泄漏声纹；管廊输气线路泄漏声纹；管廊供电线路失效与供水管路泄漏组合声纹；管廊供电线路失效与输气线路泄漏组合声纹；管廊供水管路泄漏与输气线路泄漏组合声纹）。通过模拟实际管廊的结构和尺寸比例，设计管廊数据采集实验平台，在管廊试验台中布置管线和采集设备，模拟不同工况下管线失效产生的声纹，比如：供水管道泄漏时不同的管道压力，不同的泄漏口大小和位置；供气管道泄漏时不同的管道压力，不同的泄漏口大小和位置。
所有采集的声学数据均保存为WAV格式，以保证音频数据的高保真度，避免因压缩而造成的频谱信息丢失。采集过程中严格记录每一条数据的元信息，包括声源类型、泄漏位置、泄漏尺寸、管道压力、采集距离等。试验完成后，所有音频数据将被人工标注并存储到管廊声纹数据库中。在标注阶段，还将对音频进行时间切片处理，将每段音频分割为1秒时长的文件。同时对采集到的数据需要进行人工筛选，去掉不合格的音频和音频中空白的部分，保证音频的质量。
管廊管线故障音频声纹数据集包含9个类别，分别是供水管道泄漏声纹，供气管道泄漏声纹，供电线路泄漏声纹、巡检脚步声纹、排风机声纹、供水管道泄漏与供气管道泄漏组合声纹、供水管道泄漏与供电线路泄漏组合声纹、供气管道泄漏与供电线路泄漏组合声纹、供水管道泄漏与供气管道泄漏与供电线路泄漏组合声纹，共9600条数据，293MB。

The acoustic fingerprint data of utility tunnel pipeline faults was collected in the pilot utility tunnels (Rongxi Utility Tunnel Haodan Section Comprehensive Warehouse and Telecommunication-Power Warehouse; Rongdong Utility Tunnel Section 205 Energy Warehouse, Water-Telecommunication Warehouse and High-voltage Warehouse; Rongdong Utility Tunnel Section 207 Water-Telecommunication Warehouse) and the laboratory utility tunnel test bench. In the pilot utility tunnels, soundprints of patrol footsteps and fan operation were collected using soundprint sensors and recording hosts; in the Vehicle Operation Safety Laboratory of Tangshan Research Institute, Beijing Jiaotong University, acoustic fingerprints related to utility tunnel failures were collected via the experimental platform, including power supply line failure soundprint, water supply pipeline leakage soundprint, gas transmission pipeline leakage soundprint, combined soundprint of power supply line failure and water supply pipeline leakage, combined soundprint of power supply line failure and gas transmission pipeline leakage, and combined soundprint of water supply pipeline leakage and gas transmission pipeline leakage. The utility tunnel data acquisition experimental platform was designed by simulating the actual structure and dimensional ratio of real utility tunnels. Pipelines and acquisition equipment were arranged in the test bench to simulate soundprints generated by pipeline failures under different working conditions, such as different pipeline pressures, leakage port sizes and positions during water supply pipeline leakage, as well as different pipeline pressures, leakage port sizes and positions during gas supply pipeline leakage.

All collected acoustic data are saved in WAV format to ensure high fidelity of the audio data and avoid loss of spectral information caused by compression. During the acquisition process, metadata of each data entry is strictly recorded, including sound source type, leakage location, leakage size, pipeline pressure, acquisition distance, etc. After the experiment, all audio data will be manually annotated and stored in the utility tunnel soundprint database. During the annotation stage, time slicing processing will be performed to split each audio segment into 1-second files. Meanwhile, manual screening of the collected data is required to eliminate unqualified audios and silent segments within the audios, so as to ensure the audio quality.

The utility tunnel pipeline fault audio soundprint dataset contains 9 categories, namely water supply pipeline leakage soundprint, gas supply pipeline leakage soundprint, power supply line leakage soundprint, patrol footsteps soundprint, exhaust fan soundprint, combined soundprint of water supply and gas supply pipeline leakage, combined soundprint of water supply pipeline leakage and power supply line leakage, combined soundprint of gas supply pipeline leakage and power supply line leakage, and combined soundprint of water supply, gas supply pipeline leakage and power supply line leakage. There are 9600 data entries in total, with a total size of 293 MB.

搜集汇总

背景与挑战

背景概述

该数据集包含管廊管线故障声纹数据，采集于试点管廊和实验室管廊试验台，涵盖供水管道泄漏、供气管道泄漏、供电线路泄漏等多种故障类型的声纹数据，共9600条WAV格式音频文件，总计293MB。这些数据用于管廊管线故障监测和智能监测技术研发。