Development of a new structural damage detection method using chaos attractor re-composition technique

第一个目的，我们提出了一种新的基于吸引子的结构损伤检测技术，利用混沌激励。我们从响应中重构吸引子，并通过递归分析从响应吸引子中获得定量特征。我们将此功能用作损坏指示器。通过比较完整和受损结构之间的指标，我们检测和定位结构损伤。通过数值算例表明，该方法对噪声的鲁棒性和对微小损伤的敏感性，在输入激励和观测数据均为高噪声的情况下，对结构损伤程度的识别具有良好的鲁棒性。我们发展了松弛型最小方差滤波器，改善了经典最小方差滤波器的不足。通过对一个10层模型建筑的动力参数识别，验证了该方法的有效性。

For the first purpose we propose a new attractor-based structural damage detection technique using chaotic excitation. We reconstruct attractors from responses and obtain quantitative features from the response attractors by using Recurrence Analysis. We use this feature as damage indicator. By comparing indicators between intact and damage structures, we detect and localize structural damage. The robustness against noise and sensitivity to minor damage of the proposed method are shown through numerical examples, in which we assume minor damage to a structure and high noises on input excitation and observed data.For the second purpose we apply the Monte Carlo Filter technique to identify the level of structural damage. We develop the Relaxation MCF, by which we improve the deficit of the classical MCF. The efficiency of the proposed method is confirmed by identifying the dynamic parameters of a ten-story model building be using noise contaminated simulated structural response data.