基于自适应模糊神经网络的管道剩余强度评价
Evaluation on Residual Strength of Pipeline Based on Adaptive Fuzzy Neural Network
浏览(2256) 下载(4)
- 引用格式:
-
黄亚晖,周丽丽.基于自适应模糊神经网络的管道剩余强度评价[J].天然气与石油,2011,29(6):0.doi:
.Evaluation on Residual Strength of Pipeline Based on Adaptive Fuzzy Neural Network[J].Natural Gas and Oil,2011,29(6):0.doi:
- DOI:
- 作者:
- 黄亚晖 周丽丽
- 作者单位:
- 中国市政工程中南设计研究总院
- 关键词:
- 神经网络；管道；剩余强度；评价
Neural network； Pipeline； Residual strength； Evaluation
- 摘要:
- 油气长输管道的腐蚀剩余强度评价一直是管道完整性管理的重要内容，国外对腐蚀管道的剩余强度评价取得了很大的成就。也建立了比较成熟的油气管道腐蚀评价规范。但由于油气管道腐蚀的机理复杂，腐蚀的分类很难详细界定，各种腐蚀评价标准对不同使用年限、不同等级钢材管道有其各自适应性，利用神经网络的“黑箱”原理、在不能详细了解管道腐蚀原理、不能对腐蚀进行分类的情况下，根据收集的爆破实验数据对神经网络进行模拟，通过学习好的神经网络对管道腐蚀剩余强度进行评价，与RSTRENG软件的计算结果进行了比较，取得了较好的效果。
Assessment on residual strength of corroded long-distance oil and gas pipelines has always been an important part in pipeline integrity management. Great achievement has been made in evaluation on residual strength of corroded oil and gas pipelines in foreign countries and advanced criterions have been developed for oil and gas pipeline corrosion assessment. However， it is very difficult to define classification of oil and gas pipeline corrosion accurately due to its complex mechanism and various assessment criterions for pipeline corrosion have their own self adaptability to those oil and gas pipelines which have different service life and different steel classes. In case of unidentified mechanism and classification of pipeline corrosion， the Black Box principle of neural network is applied to simulate and study the neural network system according to the experimental blasting data collected already and assess the residual strength of corroded pipelines. Compared are the assessment results with those computed by RESTRENG software and good effect is obtained.
