基于LS-SVM的压电智能结构损伤主动监测  

Active Damage Monitoring for Piezoelectric Smart Structures Based on LS-SVM Technology

作　　者：谢建宏[1] 石立华[2] 梁大开[3] 邓海[1] 

机构地区：[1]江西财经大学电子学院,江西南昌330013 [2]解放军理工大学工程兵工程学院,江苏南京210007 [3]南京航空航天大学智能材料与结构航空科技重点实验室,江苏南京210016

出　　处：《压电与声光》2007年第3期350-353,共4页Piezoelectrics & Acoustooptics

基　　金：国家自然科学基金资助项目(90205031);江西省教育厅科技基金资助项目(赣教技字[2007]277号)

摘　　要：基于被动监测技术的局限性,搭建了损伤主动监测系统,对监测信号进行了功率谱密度最大值(PSM)特征提取,并提出了一种基于最小二乘支持向量机(LS-SVM)的损伤检测方法。采用该方法,对压电智能复合材料层板进行了损伤定位的研究,并与改进的BP网络进行了对比,结果表明:在相同性能指标下,LS-SVM有比BP网络更高的损伤定位精度及更强的泛化能力。LS-SVM与主动监测技术的融合,为结构实现在线实时准确监测提供了一种新途径。Due to the limitation of passive monitoring technology, an active damage monitoring system is set up. In this system, the characteristics of monitoring signals are extracted by the method of power spectrum density maximum （PSM）, and least square support vector machine （LS-SVM） is proposed to detect damages. LS-SVM is applied to detect the damage locations for the piezoelectric smart composite laminated plates, and compared with the improved BP neural network. The results show that LS-SVM possesses the advantages such as the higher accuracy, better dissemination ability etc. under the same performance index as BP. The active monitoring technology combined with LS-SVM provides a new approach to carry out on-line, real-time, and accurate monitoring for structural damages.

关 键 词：压电智能结构 主动监测技术 功率谱密度 最小二乘支持向量机 

分 类 号：TB381[一般工业技术—材料科学与工程] TP39[自动化与计算机技术—计算机应用技术]