(b) Damage Scenario 2 FIGURE 3 F-TO-ENTER RATIOS SHOWING DAMAGED ELEMENTS AND THEIR ORDER OF SELECTION BY PARAMETER SUBSET SELECTION METHOD FIGURE 4(a), and (c) show ‘90-subspace angle’ for all the elements after the selection process. In terms of the subspace angle, damaged elements were distinguishable from other undamaged elements. FIGURE 4(b), and (d) show the extent of the damage determined by the steady state genetic algorithm for the damage scenario 1~4. Although undamaged elements are erroneously selected, the steady state genetic algorithm plays a role of both identifying incorrectly chosen elements and measuring extents of the damage. The proposed damage diagnosis methodology was confirmed to be an excellent tool combined with the modal identification for structural health monitoring through the testing under multiple severely and lightly damaged cases. SUMMARY AND CONCLUSIONS In conclusions, the proposed damage diagnosis method has been demonstrated to quantitatively assess the damage in structures under ambient environments. For future research, the proposed method should be assessed with physical experiments or in-situ ambient measurements considering other types of ambient environmental effects such as temperature variations. In the near future, the wireless smart sensor network deployed to structures will be also used to assess the performance of the proposed methodology for structural health monitoring. ACKNOWLEDGEMENT This research is supported by the New Faculty Startup Fund from the University of Akron. The material is partly based upon work supported by the National Science Foundation under Research Grant No. CMMI-0625640. Any opinions, findings, and conclusions or recommendations expressed in this paper are those of the author and do not necessarily reflect the views of the sponsors. The author is grateful to Professor Shirley J. Dyke at Washington University in St. Louis for contributions to the work supported by the National Science Foundation.
介绍 在基于振动结构健康监测的传感器技术和研究成果的最新发展使得实现在现实生活中的结构可行健康监测系统。大多数民用结构是在各种环境振动的环境中,如交通资讯,风和行人引起的振动等,在一段时间(例如,若干个周期的天数)测得的环境振动数据,具有价格低廉的优点,不需要操作干扰来激发结构,并且可以通过只输出系统识别技术可容易地用于模态识别。论文网
对于模态识别,各种技术是目前包括特征系统实现分析(ERA)技术[ Juang et al.1985 ],易卜拉欣时域法(ITD)[易卜拉欣等人.1977 ]和随机子空间迭代(SSI)技术[范overschee等人.1996 ]。最近,人们已经知道,对识别出的模态特性(固有频率和阻尼比),在一定程度上与环境振动的振幅有关,文献[ Nagayama等人的。2005,siringoringo等人.2008 ]。尽管振幅阻尼固有依赖性,还是有一些关于阻尼效应的模态识别结果以及各种识别技术的研究,[方等.1999 ]。考虑到越来越多地使用各种阻尼装置和损坏或老化结构的状态监测的重要性,研究阻尼对识别出的模态特性的影响。另一方面,环境的基于振动的损伤检测在SHM研究与实践得到了社会的广泛关注。1997,福斯维尔等人提出了一种用于定位损伤的参数子集选择方法。该方法采用特征灵敏度,测量之间的损坏和结构损坏状态的自然频率的差异[福斯维尔等人。1997 ]。云等人提出了一种基于动态剩余力测度的参数子集选择方法,可以准确识别多点损伤位置。[2008a,云等人。2008B ]。对使用模态扩展方法和噪声引入的模态识别过程中引入的噪声的不完整的动态测量方法进行了验证。然而,该基于环境振动的损伤检测的方法还没有被证明。
在本文中,一个新的健康监测策略已经提出了整合下一个时代的技术与模型为基础的损伤识别方法。它由四个主要步骤:1)测量环境振动响应;2)模态识别,使用输出的技术,下个时代的时间域;3)损伤定位使用一个新的参数子集选择方法;4)损伤量化制定优化问题。此外,为了区分真实的物理模式在未来时代技术非物理的计算模式,改进的一致性指标是在大范围内的阻尼比是一致的,具有更好的参数通过调查研究的建议。一个有限元模型的桁架桥结构已建成的数值案例研究。在环境振动下,对四种不同损伤场景进行了一系列的数值计算研究。最后,所提出的方法具有潜在的损害诊断或有限的模型更新为基础的状态评估的剩余结构寿命评估。 模态识别和结构健康监测英文文献和中文翻译(4):http://www.751com.cn/fanyi/lunwen_42972.html