【作者】 何浩祥；

【导师】 闫维明； 周锡元；

【作者基本信息】 北京工业大学 ， 结构工程， 2006， 博士

【摘要】 空间结构形式新颖丰富并且采用了大量的新材料、新技术,成为反映一个国家建筑科学技术水平的标志。空间结构所处环境状况复杂,发生损伤和破坏的潜在危险性较大,对空间结构在建设和运营期间的健康监测、诊断以及各种灾害影响下的损伤预测和识别进行研究,具有重要的理论价值和现实意义。本文结合结构健康监测和结构评估领域国内外发展现状,对空间结构健康监测问题,包括模型修正、传感器优化布置、损伤识别和信息融合等,进行了深入系统的研究,取得了如下成果:1.对小波理论在结构健康监测中基本应用的研究。对结构的动力方程进行小波灵敏度分析,得到以加速度表示的灵敏度。通过小波系数模极大值求得Lipschitz指数,将其作为衡量突变程度的指标,由此可以识别结构发生损伤的时间。不同损伤状态下的结构节点振动信号经小波包分解后在各频带上的投影是不同的,将其作为特征向量可以实现对结构的损伤程度的识别。2.对空间结构模型修正的研究。针对训练神经网络需要大量样本的情况,在结构损伤模拟的试验方案设计中,选用了均匀设计法构造样本以减少所需样本数量。用遗传算法优化BP网络的初始权值,提高神经网络的运算速度。针对空间结构的特点,提出基于子结构和神经网络的递推模型修正方法,该方法将结构分解成多层次的子结构,选取适当的损伤因素实现逐步逐级的修正。提出采用小波频带能量作为损伤因素的修正方法。3.对传感器优化布置的研究。根据结构系统可观性,给出具有重频的一般空间结构前若干阶模态所需要的最少传感器数目的估计方法。提出一种选用Frobenius范数求最大化Fisher信息矩阵的方法,在此基础上并考虑传感器优化布置的多重标准,提出基于节点能量和模态保证准则的传感器优化布置方法。针对小波智能方法的特点,运用图论理论,将空间结构或其一部份抽象为有向图;根据各杆件的连接关系确定距离矩阵,运用Floyd算法并综合考虑结构模态动能和模态变形能等参数,提出一种概念新颖的传感器优化布置方法。更多还原

【Abstract】 Large space structure are the symbol of the national building science and technology for not only its novel and prolific style but also plenty of new material and new technology have been adopted. Space structures are inevitable to suffer from environmental corrosion, long term fatigue effects or natural disasters, and then the damage accumulates during long service period. Therefore, intelligent heath monitoring and damage diagnosis for structures become an important technology to study.Detailed research are carried out for space structure heath monitoring by combining the status quo of structural health monitoring and structural condition assessment. The related important problems about structural health monitoring, such as model update, optimal placement of sensors, damage detection and information fusion, are lucubrated and a series of new methods are proposed. The main contents of this dissertation are described as follows:1. Sensitivity analysis is carried out for structural dynamic equation by wavelet, and the accelerate signal sensitivity is solved. Damage time for a structure can be detected by Lipschitz index, which is gained by module maximum of wavelet transform and is viewed as an index for identifying mutational degree of signals. Damage degree for a structure can be detected by energy spectrum at different frequency bands for the signal decomposed by wavelet packets.2. Model update for space structure is studied. Because too much samples for training Artificial Neuron Network are demanded, uniform design method is used to produce samples in order to reduce the amount of samples in the process of structural damage simulation. Genetic Algorithm is used to optimize the initial weight of back propagation network and the operation efficiency is enhanced. A recurrence update method based on substructures and ANN is presented especially for spatial structures. A structure is divided into multilayer substructures, the updating is realized step by step and a proper damage factor is更多还原