油气田瓦斯隧道TSP图像智能识别研究  

Research on Intelligent Recognition of TSP Images of Gas Tunnels in Oil and Gas Fields

作　　者：徐世权 XU Shiquan(China Railway 19th Bureau Group First Engineering Co.,Ltd.,Liaoyang 111010,China)

机构地区：[1]中铁十九局集团第一工程有限公司,辽阳111010

出　　处：《现代交通技术》2024年第3期47-50,60,共5页Modern Transportation Technology

摘　　要：在传统的非煤系油气田区瓦斯隧道超前地质预报中,由于TSP(tunnel seismic prediction,隧道地震勘探)图像的复杂性,解译的效率低且结果准确度不高,利用智能识别技术可以有效解决相应问题。依托某油气田区高瓦斯隧道工程,结合深度学习理论,借助Python(计算机编程语言)平台建立智能识别预测模型,通过该模型识别TSP图像中的P波,判断不良地质情况。结果表明:建立的智能识别模型可有效识别P波,判断围岩裂缝情况,减小人工识别的误差,实现油气田区高瓦斯隧道的地质精准探测与瓦斯防治。Due to the complexity of TSP images,the accuracy and efficiency of the interpretation results are not high in the traditional geological prediction of gas tunnels in non-coal measure oil and gas fields.The use of intelligent recognition technology can effectively solve corresponding problems.Based on the project of high gas tunnel in an oil and gas field area,the intelligent recognition and prediction model is established by using the python platform combined with the theory of deep learning.Through the model,the P-wave pattern in TSP graph is identified and the bad geological situate is judged.The results show that the intelligent recognition model can effectively identify the P-wave,judge the situation of the surrounding rock,reduce the error of manual recognition and realize precise geological detection and gas prevention of high gas tunnels in oil and gas fields.

关 键 词：智能识别 围岩裂缝 瓦斯隧道 TSP超前地质预报 

分 类 号：U456.3[建筑科学—桥梁与隧道工程]