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.
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)
关键词
支持向量机
柔度矩阵
损伤
井架
回归算法
support vector machine
flexibility matrix
damage
derrick
regression algorithm
