摘要
海底腐蚀管道极限承载力是影响管道安全的重要因素,掌握极限承载力的预测方法对于保障管道安全具有重要意义。基于广义回归神经网络的基本原理,在优化广义回归神经网络相关参数的基础上,预测了海底腐蚀管道的极限承载力。结果表明,该预测方法所得管道极限承载力的预测值和有限元计算值吻合程度较好,最大相对误差为8.34%、最小相对误差为1.92%、平均相对误差为4.67%,该方法可用于管道极限承载力的预测;不同光滑因子对预测结果的影响较大,随着光滑因子的增加,所得的均方误差增加,故在预测过程中应对光滑因子合理选择;在应用广义回归神经网络预测时,网络训练中需要调节的参数只有一个光滑因子,可达到较快的收敛速度。
The ultimate bearing capacity of the submarine corrosion pipeline is an important factor affecting the safety of the pipeline.Mastering the prediction method of ultimate bearing capacity has important practical significance for the safety of the pipeline.Based on the basic principle of generalized regression neural network,the ultimate bearing capacity of the submarine corrosion pipeline was predicted on the basis of optimizing the relevant parameters of generalized regression neural network.The results show that the predicted values of the ultimate bearing capacity of the pipeline are in good agreement with the FEM calculated values by using this prediction method,the maximum relative error is 8.34%,the minimum relative error is 1.92%,the average relative error is 4.67%,so the method can be used to predict the ultimate bearing capacity of the pipeline.Different smooth factors have great influence on the prediction results,with the increase of smooth factor,the mean square error increases,so the smooth factor should be chosen reasonably in the process of prediction.When applying the generalized regression neural network for the prediction,the parameters that need to be adjusted in network training have only one smooth factor,which can achieve faster convergence speed.
作者
靳文博
肖荣鸽
田震
李凯
JIN Wenbo;XIAO Rongge;TIAN Zhen;LI Kai(College of Petroleum Engineering,Xi'an Shiyou University,Xi'an 710065,China;Shaanxi Key Laboratory of Advanced Stimulation Technology for Oil&Gas Reservoirs,Xi'an 710065,China;OFFSHORE Oil Engineering Co.,Ltd.,Tianjin 300461,China)
出处
《热加工工艺》
北大核心
2020年第8期58-61,共4页
Hot Working Technology
关键词
广义回归神经网络
腐蚀管道
极限承载力
光滑因子
预测精度
generalized regression neural network
corrosion pipeline
ultimate bearing capacity
smooth factor
prediction accuracy
作者简介
靳文博(1986-),男,陕西咸阳人,讲师,博士,主要研究方向:油气管道流动安全保障技术,电话:18829069569,E-mail:jinwenbo725@163.com;通讯作者:肖荣鸽(1978-),女,陕西兴平人,副教授,博士,主要研究方向:油气管道流动安全保障技术,电话:13572960817,E-mail:xiaorongge@163.com。
