Damage Identification of A TLP Floating Wind Turbine by Meta-Heuristic Algorithms

Damage identification of the offshore floating wind turbine by vibration/dynamic signals is one of the important and new research fields in the Structural Health Monitoring（SHM）. In this paper a new damage identification method is proposed based on meta-heuristic algorithms using the dynamic response of the TLP（Tension-Leg Platform） floating wind turbine structure. The Genetic Algorithms（GA）, Artificial Immune System（AIS）, Particle Swarm Optimization（PSO）, and Artificial Bee Colony（ABC） are chosen for minimizing the object function, defined properly for damage identification purpose. In addition to studying the capability of mentioned algorithms in correctly identifying the damage, the effect of the response type on the results of identification is studied. Also, the results of proposed damage identification are investigated with considering possible uncertainties of the structure. Finally, for evaluating the proposed method in real condition, a 1/100 scaled experimental setup of TLP Floating Wind Turbine（TLPFWT） is provided in a laboratory scale and the proposed damage identification method is applied to the scaled turbine.

Damage Identification under Incomplete Mode Shape Data Using Optimization Technique Based on Generalized Flexibility Matrix

A generalized flexibility–based objective function utilized for structure damage identification is constructed for solving the constrained nonlinear least squares optimized problem. To begin with, the generalized flexibility matrix (GFM) proposed to solve the damage identification problem is recalled and a modal expansion method is introduced. Next, the objective function for iterative optimization process based on the GFM is formulated, and the Trust-Region algorithm is utilized to obtain the solution of the optimization problem for multiple damage cases. And then for computing the objective function gradient, the sensitivity analysis regarding design variables is derived. In addition, due to the spatial incompleteness, the influence of stiffness reduction and incomplete modal measurement data is discussed by means of two numerical examples with several damage cases. Finally, based on the computational results, it is evident that the presented approach provides good validity and reliability for the large and complicated engineering structures.

HOMOTOPY CONTINUATION ALGORITHM FOR STRUCTURAL DAMAGE IDENTIFICATION

A new method is put forward for structural damage identification based on the homotopy continuation algorithm. A numerical example is presented to verify the method. The beams with different damage locations and different damage extents are identified by this method. The numerical examples have proved that this new method is capable of easy convergence, which is not sensitive to the initial iterative values. It is effective for accurately identifying multiple damages. By incorporating the finite element method into the homotopy continuation algorithm, the damage identifying ability of the new method can be greatly enhanced.

Three-Step Damage Identification Method Based on Dynamic Characteristics

A three-step damage identification method based on dynamic characteristics is proposed to improve the structure reliability and security and avoid serious accident. In the proposed method, the frequency and difference of modal curvature(DMC) are used as damage indexes. Firstly, the detection of the occurrence of damage is addressed by the frequency or the square of frequency change. Then the damage location inside the structure is measured by the DMC. Finally, with the stiffness reduction rate as a damage factor, the amount of damage is estimated by the optimization algorithm. The three-step damage identification method has been validated by conducting the simulation on a cantilever beam and the shaking table test on a submerged bridge. The results show that the method proposed in this paper can effectively solve the damage identification problem in theory and engineering practice.

Damage Identification in Beam-Type Structures Using Pseudo Strain Energy Density and Grey Relation Coefficient

Based on pseudo strain energy density (PSED) and grey relation coefficient (GRC), an index is proposed to locate the damage of beam-type structures in time-domain. The genetic algorithm (GA) is utilized to identify the structural damage severity of confirmed damaged locations. Furthermore, a systematic damage identification program based on GA is developed on MATLAB platform. ANSYS is employed to conduct the finite element analysis of complicated civil engineering structures, which is embedded with interface technique. The two-step damage identification is verified by a finite element model of Xinxingtang Highway Bridge and a laboratory beam model based on polyvinylidens fluoride (PVDF). The bridge model was constructed with 57 girder segments, and simulated with 58 measurement points. The damaged segments were located accurately by GRC index regardless of damage extents and noise levels. With stiffness reduction factors of detected segments as variables, the GA program evolved for 150 generations in 6 h and identified the damage extent with the maximum errors of 1% and 3% corresponding to the noise to signal ratios of 0 and 5%, respectively. In contrast, the common GA-based method without using GRC index evolved for 600 generations in 24 h, but failed to obtain satisfactory results. In the laboratory test, PVDF patches were used as dynamic strain sensors, and the damage locations were identified due to the fact that GRC indexes of points near damaged elements were smaller than 0.6 while those of others were larger than 0.6. The GA-based damage quantification was also consistent with the value of crack depth in the beam model.

An Optimized Damage Identification Method of Beam Using Wavelet and Neural Network

An optimized damage identification method of beam combined wavelet with neural network is presented in an attempt to improve the calculation iterative speed and accuracy damage identification. The mathematical model is developed to identify the structure damage based on the theory of finite elements and rotation modal parameters. The model is integrated with BP neural network optimization approach which utilizes the Genetic algorithm optimization method. The structural rotation modal parameters are performed with the continuous wavelet transform through the Mexico hat wavelet. The location of structure damage is identified by the maximum of wavelet coefficients. Then, the multi-scale wavelet coefficients modulus maxima are used as the inputs of the BP neural network, and through training and updating the optimal weight and threshold value to obtain the ideal output which is used to describe the degree of structural damage. The obtained results demonstrate the effectiveness of the proposed approach in simultaneously improving the structural damage identification precision including the damage locating and severity.

结合蜉蝣算法和响应面的结构静力损伤识别

提出了一种结合蜉蝣智能搜索算法和静位移响应面的结构损伤识别方法.首先构造了基于灵敏度的静力损伤识别方程残差和位移残差指标的正则化优化目标函数,然后采用人工智能领域的蜉蝣算法来识别结构单元级别的损伤.在优化过程中,为了解决静力加载点和位移测点不一致的问题,利用模型凝聚和位移扩展方法重新构建了损伤识别方程.同时,利用静位移响应面计算目标函数中的位移残差,避免了耗时的有限元计算,提高了优化效率.简支梁的数值算例结果表明,本文方法在寻优速度和精度方面均优于传统的粒子群法和差分进化算法;与单独基于静力损伤识别方程残差或位移残差指标的方法相比,识别损伤更为准确.铝合金悬臂梁损伤识别的静载实验进一步验证了本文方法的高效率和有效性.

基于融合WM算法的Mamdani型模糊推理的板桥损伤定量分析

为对板桥局部位置的损伤程度进行定量分析,提出1种基于融合WM算法的Mamdani型模糊推理板桥损伤定量分析方法,探讨影响面指标和损伤程度之间的独特联系,并通过MIDAS及Python模拟计算并分析板的损伤程度。研究结果表明:模糊逻辑系统具有较为出色的记忆能力。当融合WM算法到模糊逻辑系统中时,不论是单点损伤还是多点损伤,都能对板桥的损伤程度进行准确识别;当未融合WM算法时,由于其模糊规则没有进行置信度计算,将错误的规则也计算在内,最终导致定量结果精度下降。研究结果可为板桥结构的损伤评估和维护策略提供较为科学准确的参考。

基于遥感数据的工程结构损伤动态识别研究

以定期监测工程结构状况为目标,提出了基于遥感数据的工程结构损伤动态识别方法,解决获取影像粗糙问题,提升微小工程结构损伤识别效果。运用二维Tsallis交叉熵计算工程结构遥感影像的二维直方图,通过在布谷鸟搜索算法(CS)中引入Logistic映射的混沌扰动算子,形成混沌CS算法,完善二维Tsallis交叉熵的阈值选择过程,获取最佳阈值后,分割工程结构遥感影像,增强处理分割后的遥感影像,并将处理后的影像作为SVM分类识别模型输入,完成损伤动态识别、归类。实验结果表明,该方法获取遥感影像精度极高,分割影像清晰且各区域边缘完整,能够识别微小的裂缝损伤,最大程度还原裂缝宽度及线性特征,识别结果影像信息含量大,且识别均方误差低、平方相关系数高。

两阶段注采管损伤仿真识别方法研究

注采管的微裂缝或微损伤会逐渐发展为疲劳断裂破坏,给储气库的安全运营带来巨大挑战。于是,以注采管为对象,将改进灰关联度与柔度曲率差相结合形成新的损伤识别指标,提取出流固耦合作用下注采管有限元模型的固有频率和振型,以计算改进灰关联柔度曲率差指标(IGMFC)值,对结构损伤进行定位,并利用改进后的粒子群算法(IPSO)优化极限学习机(ELM)对结构进行损伤定量。结果表明指标IGMFC具有很好的准确度、灵敏度和抗噪性,且优化后的极限学习机量化损伤程度的误差在2%以内,损伤识别结果误差仅为2.92%,有一定的应用价值。

基于改进灰狼优化算法的结构损伤识别

结合模态柔度矩阵、广义模态柔度矩阵和振型三个识别精度较好的指标,构造新的目标函数求解损伤识别问题。通过Nelson方法求解得到的频率与振型的导数,得到对结构刚度发生变化时更具敏感性的位置,然后在这些位置布置传感器以提取结构信息。针对原有的灰狼算法虽然全局搜索能力强,但是存在局部搜索精度差的问题,本文从初始种群和收敛因子等方面着手,改善灰狼算法的局部搜索能力及收敛速度。最后利用提出的方法,通过识别梁模型及桁架模型中的损伤单元说明本文方法的有效性。

基于改进粒子群算法的层合板损伤定量检测方法

针对层合板在加工和服役过程中出现的结构损伤问题,提出基于差分法和改进粒子群算法的两阶段损伤定量检测方法。首先,采用单元弹性模量降低的方法定义了结构损伤单元,通过结构振动特性分析获得了不同损伤工况层合板数值模型的模态参数,运用差分法识别了损伤大致区域。其次,利用改进的粒子群算法定量计算损伤的精确位置和程度。所提算法通过待测结构模型和迭代更新后有限元模型模态参数的柔度矩阵构建目标函数,引入变异因子以使算法有效的跳出局部收敛,MS算子减少了健康单元对真实损伤单元损伤程度的影响。结果表明:基于改进的粒子群算法可精确识别复合材料层合板结构损伤单元的位置和损伤程度。

基于遗传算法的高速铁路桥梁损伤识别

针对高速铁路桥梁结构服役局部损伤问题,文章提出一种改进损伤识别因子并结合遗传算法的损伤识别方法。推导了任意单元长度组合的车辆-轨道-桥梁耦合单元及其动力方程,遗传算法采用锦标赛策略及最优保存策略2种选择算子,引入灾变及邻域搜索机制。通过对不同损伤工况进行多次独立重复计算,对算法的稳定性与准确性进行统计分析。结果表明:在信噪比(signal-to-noise ratio,SNR)大于25 dB时,识别效果好,极少数误判的结果大概率存在于真实损伤位置的附近或对称处;与单个单元损伤情况相比,多个单元损伤工况平均第一次出现最优解的迭代次数略高,但整体都偏低;SNR与识别率(identification rate,IR)呈“扁S”型Logistic曲线关系。

基于超声检测的主动式钢轨伤损智能识别方法

常规的主动式钢轨伤损智能识别方法,在钢轨伤损数据采集过程中耗时较多,使得伤损智能识别时间较长。为解决这一问题,提出一种基于超声检测的主动式钢轨伤损智能识别方法。首先按照增益范围,获取钢轨伤损回波信号,设置超声换能器的滤波器,对信号进行转换,从而对数据进行采集;然后根据DBSCAN算法,以伤损识别的最小单元为一个单元,对采集的超声数据进行组合而构成超声信息群,并对数据进行划分,按照划分数据对钢轨伤损特征进行匹配;最后以AlexNet网络架构为识别主体结构,以数据匹配结果为基础,建立显图样本数据集,对伤损类型进行精细搜索,从而得到主动式钢轨伤损智能识别结果。试验结果表明,所提方法对主动式钢轨伤损识别的时间较短,能实现对主动式钢轨伤损的快速识别,具有较好的应用价值。

基于模型修正与灰色系统理论的梁桥损伤识别研究

对既有梁桥结构开展损伤识别分析时,需要依托有限元软件建立理想的梁桥数值模型,然而依据设计图纸建立的梁桥有限元模型尚未考虑到建模过程的不确定性,有限元模型计算与既有梁桥实测存在较大偏差。利用基于遗传算法优化BP神经网络的模型修正方法,对修正后更加贴近实际梁桥结构的有限元模型展开数值分析,通过挠度影响线进行灰色系统相关性分析与损伤识别。修正结果表明:基于遗传算法优化BP神经网络的梁桥模型修正方法可使有限元模型贴近实际结构模型;结合现场模态测试,结果频率误差从未修正的30.5%降至修正后的9.8%。损伤识别结果表明灰色系统理论相关系数与损伤程度呈线性关系,灰色系统相关系数越小,损伤程度越大。

基于樽海鞘算法优化支持向量机的连续梁桥损伤识别

为了能够更加准确、高效地判断桥梁结构损伤位置和程度,本文提出了基于樽海鞘群优化支持向量机(SSA-SVM)方法进行连续梁桥损伤识别的方法。该方法以敏感性较高的曲率模态差作为损伤识别指标,利用樽海鞘群(SSA)算法寻找支持向量机(SVM)最优参数,建立SVM预测模型,通过建立一座三跨连续梁桥有限元模型,以桥梁易损区域作为损伤识别对象进行数值模拟。结果表明:以曲率模态差作为损伤识别指标,能够有效地识别并定位单点或多点的损伤状况,同时准确评估损伤的严重程度。与传统SVM模型比较,SSA-SVM模型实现了参数的自动优化,同时也拥有了更为精准的预测能力。

基于改进YOLOv8的风电叶片表面损伤检测与识别方法

针对风电叶片极易出现损伤和故障,且制造和维护成本高昂等问题,提出了一种基于改进YOLOv8模型的风电叶片表面损伤检测与识别方法。首先,将现场拍摄到的高清叶片图像作为实验数据集,并将其按比例随机划分为训练集、验证集和测试集;然后,在YOLOv8模型中引入了动态数据增强算法Mosaic、Mixup及离线数据增强算法Albumentations,对训练数据集进行了扩充,解决了模型在有限数据集下的泛化性问题;最后,使用卷积注意力模块(CBAM)和梯度协调机制(GHM)/Focal loss算法等手段加强了模型的损伤检测能力,改进了样本分布不均衡问题,建立了一种先进的风电叶片表面损伤检测与识别方法,提升了YOLOv8模型对叶片损伤的检测精度。研究结果表明:改进后的YOLOv8模型在计算量和参数量都较低的情况下,其平均精度(AP)、平均召回率(AR)都超越了同等配置下的快速区域卷积神经网络(Faster R-CNN)模型。改进后的YOLOv8模型在交并比(IoU)阈值为0.5时的AP和AR分别达到了73.2%和58.8%,验证了该方法在风电叶片损伤检测方面具有一定的可靠性和有效性。

利用模态应变能变化率的结构损伤识别优化方法

针对结构损伤识别过程中存在的定位精度和量化分析不足的问题,提出一种基于结构振动响应的模态应变能变化率与优化技术相结合的损伤识别方法。利用两步法确定可疑损伤单元及对其损伤程度进行量化分析。通过有限元法建立结构的损伤特征模型,且利用单元模态应变能变化率指标构建损伤指标优化分析的目标函数。数值分析过程中,利用粒子群优化算法与遗传算法对设计变量进行优化分析。同时比较模态应变能变化率与小波分析两种方法下的损伤定位的量化分析效果和识别效率。在实际算例中,利用梁结构进行损伤识别优化方法的结果验证。结果证明该方法能够显著提高结构振动损伤和定位的有效性。该方法不仅能够快速和精准地进行结构损伤量化分析,比小波方法具有更好的定位效果,而且能够提高量化分析的识别效率。但该方法在实施过程中的定位精度容易受到噪声的影响,通过优化目标函数可在一定程度上提高该结构损伤识别方法的抗噪能力。

基于BA-Elman算法的预应力钢筋混凝土梁损伤识别研究

为了准确识别预应力混凝土结构的损伤程度,制作预应力钢筋混凝土实验梁,进行三点弯曲加载实验,收集损伤全过程声发射(AE)信号.绘制声发射振铃计数与持续时间的特征参数关联分布图,以揭示梁的损伤演化过程.借鉴加卸载响应比理论进一步将梁的损伤破坏过程划分为4个典型阶段.构建Elman神经网络,基于Elman神经网络采用局部搜索算法,难以达到全局最优的缺点,提出用蝙蝠算法(BA)对其进行优化.设计BA-Elman神经网络模型训练识别试验梁各损伤阶段AE信号特征参数数据,准确率达到93%,相较于基础Elman神经网络准确率提高了6%左右.定型BA-Elman网络结构并识别同种工况下的其他梁AE信号,识别准确率达到92%左右.

布谷鸟算法优化神经网络的站房结构损伤识别

研究目的:在铁路站房结构健康监测中,采用BP神经网络来识别结构损伤常常因为陷入局部极小值导致误报率高。因此,本文使用布谷鸟算法对神经网络进行优化,并将优化后的神经网络模型应用于站房结构损伤识别,以提高损伤识别的准确率。研究结论:(1)利用布谷鸟算法搜索BP神经网络最优的权重和偏置,可以找到模型的全局最优解,防止陷入局部极小值;(2)本文提出的模型适用于铁路站房结构健康监测数据量大、结构复杂、维数高的特点,可通过数据预处理构建神经网络的训练测试数据集;(3)以昆明南站监测数据为例进行试验验证,结果表明,通过布谷鸟算法优化BP神经网络模型准确率达到99.73%,较BP神经网络模型提高1.88个百分点;(4)本文研究可为实时结构性能评估提供技术支撑,可为铁路站房结构健康监测提供新的途径,具有一定的参考价值。