基于小波和神经网络的石油井架损伤定位研究

详细信息 本馆镜像全文    |  推荐本文 | | 获取馆网全文

英文篇名：Study on Damage Location of Oil Derrick Based on Wavelet and Neural Network 中文刊名：科学技术与工程 英文刊名：Science Technology and Engineering 作者：邹龙庆 ; 付海龙 ; 谢春强 英文作者：ZOU Long-qing ; FU Hai-long ; XIE Chun-qiang(School of Mechanical Science and Engineering ; Daqing Petroleum Institute ; Daqing 163318 ; P.R.China) 中文关键词：石油井架 ; 小波分析 ; 神经网络 ; 损伤 英文关键词：oil derrick wavelet analysis neural network damage 出版日期：2009-02-15 机构：大庆石油学院机械科学与工程学院; 年：2009 期：04 出版单位：科学技术与工程

摘要

针对石油井架单一损伤位置识别问题,结合小波和神经网络分析方法,建立两层BP神经网络,将小波包能量分析得到的归一化能量特征向量作为网络的输入向量,经过网络训练和仿真测试,证明其在识别单一损伤位置时,具有较好的实用性。同时对石油井架相似模型进行了锤击振动试验,说明基于归一化能量特征向量的损伤评价指标可以准确实现石油井架结构损伤位置的定位识别。
According to the problem of single damage location of oil derrick,the two-layer BP neural network is build by the combination of wavelet and neural network method.The eigenvectors of normalized energy are as the neural network's inputing vectors,which are obtained from the wavelet analysis.After network training and simulation test,it is concluded that the method above has better practicability to identify single damage location of oil derrick.The vibration experiment beating the indoor derrick model with a hammer is completed,which shows that the damage evaluation guide line based on eigenvectors of normalized energy can achieve to identify damage location of oil derrick well and truly.

引文

1杨建国.小波分析及其工程应用.北京:机械工业出版社,2005
    2 Lee Y Y,Liew K M.Detection of damage locations in a beam usingthe wavelet analysis.International Journal of Structural Stability andDynamics,2001;1(3):455—465
    3吴金志,张毅刚.基于BP神经网络的网格结构损伤识别与试验研究.世界地震工程,2005;21(4):71—76
    4朱炜.大跨距K型伸缩井架及底座试验模型设计.石油机械,1995;23(1):1—6
    5丁幼亮,李爱群.基于振动测试与小波包分析的结构损伤预警.力学学报,2006;38(5):639—644