基于柔度矩阵和支持向量机的井架损伤识别技术  

Damage Identification Technique of Oil Derrick Based on Support Vector Machine and Flexibility Matrix

作　　者：邹龙庆[1] 陈桂娟[1] 付海龙[1] 赵海洋[1] 

机构地区：[1]大庆石油学院机械科学与工程学院,黑龙江大庆163318

出　　处：《石油矿场机械》2008年第2期1-4,共4页Oil Field Equipment

基　　金：黑龙江省自然科学基金资助项目(E200509)

摘　　要：与频率和振型相比,柔度矩阵是结构损伤检测的有力工具。支持向量机是在统计学理论上发展出的一种学习机器,具有优秀的回归特性。以柔度矩阵为检测指标,利用支持向量机的回归算法,提出了一套基于支持向量机和柔度矩阵的损伤识别技术,并对JJ160/41-K型石油井架进行了损伤识别和仿真计算,结果表明该技术可有效识别结构损伤的程度及位置。Study indicates that flexibility matrix is a kind of powerful method of structure damage detection. Support vector machine （SVM） is a study machine developing from statical learning theory, which has splendid regression feature. This paper puts forth a set of damage identification technique based on SVM, making flexibility matrix as testing indexs and using SVM regression algorithm. Meanwhile, validation is carried by emulational calculation of JJ160/41-K derrick, and then the results indicate that this technique can identify damage degree and location effectively.

关 键 词：支持向量机 柔度矩阵 损伤 井架 回归算法 

分 类 号：TE923[石油与天然气工程—石油机械设备]