基于范数归一化和稀疏正则化约束的结构损伤检测
|
摘要
利用结构损伤空间稀疏性进行结构损伤识别是结构健康监测领域的研究热点之一,基于结构灵敏度分析与稀疏正则化的损伤识别方法能有效识别结构损伤位置和程度,但在噪声等因素影响下,其识别结果容易出现误判和刚度强化等问题。针对此问题,基于范数归一化与稀疏正则化约束,提出了一种结构损伤检测方法;该方法在迭代求解过程中,通过增加范数归一化、稀疏正则化约束以及对模型增加牛顿迭代法约束和总损伤折减系数约束,达到减少误判、增加结果合理化和提高识别精度的目的。三种不同结构损伤识别仿真算例研究表明:增加范数归一化和模型约束后,结构损伤识别精度得到明显提高;在不同噪声水平下,所提新方法既能有效定位结构损伤又能准确识别结构损伤程度,且具有较强鲁棒性。
Using the space sparsity of structural damage is prevalent to identify structural damages in the field of structural health monitoring. Based on the sparse regularization,first-order sensitivity analysis can detect damage locations and quantify damage extents effectively. However,the misjudgments and stiffness hardening would occur under the influence of noises. A new structural damage detection( SDD) algorithm,based on the norm normalization and sparse regularization constraints,was proposed to solve these problems. It can reduce misjudgments,make the damage detection results more rational and improve identification accuracy by adding the norm normalization and sparse regularization constraints to the process of iteration as well as adding,the constraint of Newton iteration method and the total damage reduction factor to the model. The numerical simulation results from three different structures indicate that the damage detection identification are obviously improved after adding the norm normalization and model constraints. The new SDD method can effectively identify damage locations and extents under different level noises and get high robustness to noises.
Using the space sparsity of structural damage is prevalent to identify structural damages in the field of structural health monitoring. Based on the sparse regularization,first-order sensitivity analysis can detect damage locations and quantify damage extents effectively. However,the misjudgments and stiffness hardening would occur under the influence of noises. A new structural damage detection( SDD) algorithm,based on the norm normalization and sparse regularization constraints,was proposed to solve these problems. It can reduce misjudgments,make the damage detection results more rational and improve identification accuracy by adding the norm normalization and sparse regularization constraints to the process of iteration as well as adding,the constraint of Newton iteration method and the total damage reduction factor to the model. The numerical simulation results from three different structures indicate that the damage detection identification are obviously improved after adding the norm normalization and model constraints. The new SDD method can effectively identify damage locations and extents under different level noises and get high robustness to noises.
引文
[1]GERIST S,MAHERI M R.Multi-stage approach for structural damage detection problem using basis pursuit and particle swarm optimization[J].Journal of Sound&Vibration,2016,384:210-226.
[2]JAYAWARDHANA M,ZHU X,LIYANAPATHIRANA R,et al.Compressive sensing for efficient health monitoring and effective damage detection of structures[J].Mechanical Systems&Signal Processing,2017,84:414-430.
[3]MASCAREAS D,CATTANEO A,THEILER J,et al.Compressed sensing techniques for detecting damage in structures[J].Structural Health Monitoring,2013,12(4):325-338.
[4]ZHOU X Q,XIA Y,WENG S.L1 regularization approach to structural damage detection using frequency data[J].Structural Health Monitoring,2015,14(6):571-582.
[5]叶肖伟,姜洋,倪一清,等.基于FBG反射谱特征的铁路道岔损伤识别试验研究[J].振动与冲击,2014,33(6):71-76.YE Xiaowei,JIANG Yang,NI Yiqing,et al.Experimental study on damage detection of railway turnouts based on characteristics of FBG reflective spectra[J].Journal of Vibration&Shock,2014,33(6):71-76.
[6]郭冬阳.基于灵敏度和遗传方法的结构动力模型修正技术研究及系统实现[D].武汉:华中科技大学,2014.
[7]周述美,鲍跃全,李惠.基于结构灵敏度分析与稀疏约束优化的结构损伤识别方法[J].振动与冲击,2016,35(9):135-140.ZHOU Shumei,BAO Yuequan,LI Hui.Structural damage identification based on structural sensitivity analysis and sparse restrains optimization[J].Journal of Vibration&Shock,2016,35(9):135-140.
[8]TIBSHIRANI R.Regression shrinkage and selection via the lasso[J].Journal of the Royal Statistical Society,1996,58(1):267-288.
[9]LIU J,JI S,YE J.SLEP:sparse learning with efficien projections[J].Arizona State University,2009,6:491.
[10]BURNHAM K P,ANDERSON D R.Model selection and multimodel inference:a practical information-theoretic approach[M].New York:Springer Science&Business Media,2003.
[11]李成,余岭.基于人工鱼群方法的结构模型修正与损伤检测[J].振动与冲击,2014,33(2):112-116.LI Cheng,YU Ling.Structural model updating and damage detection based on artificial fish swarm algorithm[J].Journal of Vibration&Shock,2014,33(2):112-116.
[2]JAYAWARDHANA M,ZHU X,LIYANAPATHIRANA R,et al.Compressive sensing for efficient health monitoring and effective damage detection of structures[J].Mechanical Systems&Signal Processing,2017,84:414-430.
[3]MASCAREAS D,CATTANEO A,THEILER J,et al.Compressed sensing techniques for detecting damage in structures[J].Structural Health Monitoring,2013,12(4):325-338.
[4]ZHOU X Q,XIA Y,WENG S.L1 regularization approach to structural damage detection using frequency data[J].Structural Health Monitoring,2015,14(6):571-582.
[5]叶肖伟,姜洋,倪一清,等.基于FBG反射谱特征的铁路道岔损伤识别试验研究[J].振动与冲击,2014,33(6):71-76.YE Xiaowei,JIANG Yang,NI Yiqing,et al.Experimental study on damage detection of railway turnouts based on characteristics of FBG reflective spectra[J].Journal of Vibration&Shock,2014,33(6):71-76.
[6]郭冬阳.基于灵敏度和遗传方法的结构动力模型修正技术研究及系统实现[D].武汉:华中科技大学,2014.
[7]周述美,鲍跃全,李惠.基于结构灵敏度分析与稀疏约束优化的结构损伤识别方法[J].振动与冲击,2016,35(9):135-140.ZHOU Shumei,BAO Yuequan,LI Hui.Structural damage identification based on structural sensitivity analysis and sparse restrains optimization[J].Journal of Vibration&Shock,2016,35(9):135-140.
[8]TIBSHIRANI R.Regression shrinkage and selection via the lasso[J].Journal of the Royal Statistical Society,1996,58(1):267-288.
[9]LIU J,JI S,YE J.SLEP:sparse learning with efficien projections[J].Arizona State University,2009,6:491.
[10]BURNHAM K P,ANDERSON D R.Model selection and multimodel inference:a practical information-theoretic approach[M].New York:Springer Science&Business Media,2003.
[11]李成,余岭.基于人工鱼群方法的结构模型修正与损伤检测[J].振动与冲击,2014,33(2):112-116.LI Cheng,YU Ling.Structural model updating and damage detection based on artificial fish swarm algorithm[J].Journal of Vibration&Shock,2014,33(2):112-116.