Scholarlink

<p>In the Central Sichuan region, oil-gas fields are widely distributed, and due to the toxicity and explosiveness of natural gas, tunnel construction poses safety hazards. Existing prediction methods for oil-gas tunnels have deficiencies in accuracy and applicability. Therefore, developing a new method for the classification prediction of harmful gas is crucial for the design and construction of projects in the Central Sichuan region. This research proposed a prediction method based on the Support Vector Machine (SVM) optimized using the Energy Valley Optimizer (EVO) algorithm. Initially, 114 sets of harmful gas tunnel cases were selected based on existing engineering data. Parameters such as tunnel depth, length, oil-gas field location score, structural score, and lithology score were used as inputs, while the actual gas classification served as the output for model validation. The results show that the method achieved a high accuracy of 96.67%, along with higher convergence speed and higher predictive accuracy. The method was applied to the Sichuan Basin region to predict the hazard levels of 89 tunnels within the area, obtaining the hazard levels and distribution map of harmful gas in the area, providing valuable guidance for tunnel construction projects in the area.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Research on gas tunnel prediction in Central Sichuan using energy valley optimizer and support vector machine

  • 期刊论文

  • 发表日期：

    2025年1月9日

查看全文
  • Yuxuan Liu,
  • Peidong Su,
  • Peng Qiu,
  • Tao Luo,
  • Can Yang,
  • Xinghao Lu

摘要

In the Central Sichuan region, oil-gas fields are widely distributed, and due to the toxicity and explosiveness of natural gas, tunnel construction poses safety hazards. Existing prediction methods for oil-gas tunnels have deficiencies in accuracy and applicability. Therefore, developing a new method for the classification prediction of harmful gas is crucial for the design and construction of projects in the Central Sichuan region. This research proposed a prediction method based on the Support Vector Machine (SVM) optimized using the Energy Valley Optimizer (EVO) algorithm. Initially, 114 sets of harmful gas tunnel cases were selected based on existing engineering data. Parameters such as tunnel depth, length, oil-gas field location score, structural score, and lithology score were used as inputs, while the actual gas classification served as the output for model validation. The results show that the method achieved a high accuracy of 96.67%, along with higher convergence speed and higher predictive accuracy. The method was applied to the Sichuan Basin region to predict the hazard levels of 89 tunnels within the area, obtaining the hazard levels and distribution map of harmful gas in the area, providing valuable guidance for tunnel construction projects in the area.