波流联合作用下埕岛海域海床液化与管道局部冲刷预测模型

Predictive Model of Seabed Liquefaction and Local Scouring of Pipelines in the Chengdao Sea Area Under Combined Wave and Current Action

  • 摘要: 波流联合作用下的海床液化与局部冲刷,对海底管道的运行安全会构成极端不利的威胁。如何根据波流预报信息快速预测海底管道路由处的海床液化与局部冲刷,一直是海上油气管道运行管理部门关注的问题。根据埕岛海域工程地质环境背景,针对波-流-管道耦合作用下的海床液化与局部冲刷问题分别构建了对应的数值模型,开展了不同波流、管道,以及沉积物特征参数等组合工况下的模拟计算,分析得到了埕岛油田海底管道局部冲刷与海床液化的典型响应特征。在此基础上,将数值模型的关键输入与输出数据集进行归一化处理,采用机器学习模型构建了针对埕岛海域海底管道路由易致灾段的海床液化与局部冲刷快速响应分析模型,并编制了可视化程序。研究结果表明开发的预测分析软件具有较好的预测精度,且计算响应时间缩短至秒级,可以为海底管道运行安全应急管理提供一定的参考。

     

    Abstract: The liquefaction and local erosion of the seabed under the combined action of waves and currents can pose extremely adverse effects on the safety of submarine pipelines. How to quickly predict seabed liquefaction and local erosion around submarine pipelines under forthcoming waves and currents has always been a key issue for the operation and management departments of offshore oil and gas pipelines. The corresponding numerical calculation models for the seabed liquefaction and local erosion induced by the combined action of waves and currents around the submarine pipeline was established based on the engineering geological background of Chengdao sea area. Numerical calculations with different combinations of wave currents, pipelines, and sediment characteristic parameters were carried out and the typical response characteristic parameters of local erosion and seabed liquefaction were analyzed. On this basis, the key input and output parameters of the numerical examples were chosen for normalization, and were used to construct a machine learning model for seabed liquefaction and local erosion prediction in the vulnerable section of the submarine pipeline routing in the Chengdao area. The established machine learning model which has been developed into a visualization software shows good prediction accuracy, and the calculation can be finished in a few seconds. The research results can provide reference for emergency management of submarine pipelines in Chengdao sea area.

     

/

返回文章
返回