Predictive Analysis of Fluid-Hammer Effect on LNG Regasification System Pipeline Network
Predictive Analysis of Fluid-Hammer Effect on LNG Regasification System Pipeline Network
复制标题
DOI：
10.1109/rams51457.2022.9894025
复制
发表时间：
2022-01
期刊：
影响因子：
--
通讯作者：
Ajinkya Zalkikar;Bimal P. Nepal;H. Husin;O. Yadav;A. Banerjee
中科院分区：
文献类型：
--
作者：
Ajinkya Zalkikar;Bimal P. Nepal;H. Husin;O. Yadav;A. Banerjee
本文提出了各种机器学习模型的比较用于流体锤压力波动脆弱性评估的海水管道在液化天然气再气化厂。液化天然气（LNG）再气化系统是天然气生产系统中最关键的部件之一，其使用携带海水的复杂管道网络将LNG从液相转化回气相（天然气），以从LNG中提取冷能，该LNG容易形成流体锤。本文提出了一种基于机器学习的方法来预测LNG再气化系统复杂管网中流体流动的流体锤效应。该方法包括两个部分：第一，这项研究包括一个基于模拟的模型，在白杨HYESTIC开发的LNG再气化系统管道的设计参数影响的水击效应的实验设计原则的基础上，第二，各种机器学习算法，提出了预测管道的脆弱性，由于水击效应的管道。结果表明，径向核支持向量机算法是预测管道脆弱性的最佳模型。
This paper proposes a comparison of various machine learning models used for fluid hammer pressure surge vulnerability assessment in seawater pipeline in the LNG regasification plant. One of the most critical components of natural gas production system is the Liquefied Natural Gas (LNG) regasification system which converts the LNG back from liquid phase to gaseous phase (Natural Gas) using a complex pipeline network carrying seawater to extract cold energy from the LNG which is susceptible to fluid hammer formation. In this paper, machine learning based methodology is presented to predict fluid hammer effect in the fluid flow in the complex pipeline network of LNG regasification system. The methodology consists of two parts: first, this study includes a simulation-based model developed in Aspen HYSYS to find the design parameters affecting the water hammer effect in the LNG regasification system pipeline based on the Design of Experiments principles and second, various machine learning algorithms are proposed to predict the vulnerability of the pipeline due to water hammer effect in the pipeline. The Support Vector Machine algorithm with radial kernel was found to be the best model to predict the vulnerability of the pipeline.