工业技术与职业教育2023，Vol.21Issue(6)：16-19,4.
基于机器学习的深海能源土降压开采沉降预测
Prediction of Subsidence of Deep Sea Energy Soil Depressurization Mining Based on Machine Learning
摘要
Abstract
The ocean subsidence induced by gas hydrate exploitation is an important problem that must be faced in the ocean development.It is particularly important to predict the subsidence of deep-sea energy soil under the condition of depressurized exploitation.When predicting the seabed subsidence caused by gas hydrate extraction,the prediction methods such as recurrent neural network cannot fully consider the influence of uncertain factors such as temperature and pore pressure,so the prediction accuracy is low and the error is large.A short-time memory network method based on the optimization of main parameters is proposed to predict the subsidence of deep-sea energy soil depressurized exploitation,and the observation data of seabed subsidence caused by depressurized exploitation of gas hydrate in Shenhu Sea area of the South China Sea are compared.The results show that:There is a good agreement between the actual measured value and the predicted value,and the model error can meet the expected accuracy,which can be applied to the settlement prediction of deep-sea energy soil depressurization mining.关键词
深海能源土/沉降预测/机器学习/长短时记忆网络Key words
deep-sea methane hydrate soil/settlement prediction/machine learning/long and short term memory network分类
能源科技引用本文复制引用
靳继凯,温欣,张艺博,赵春晖..基于机器学习的深海能源土降压开采沉降预测[J].工业技术与职业教育,2023,21(6):16-19,4.基金项目
河北省高层次人才资助项目"开采扰动条件下深海能源土性能演变规律及多功能试验系统的开发"(项目编号:C20221066),主持人温欣. （项目编号:C20221066）
