STUDY ON THE DAMAGE DETECTION OF SPACE STRUCTURE SYSTEMS BY SYSTEM IDENTIFICATION APPROACH

采用小尺度试验模型对空间桁架和框架结构进行了损伤检测。这些结构由大量的成员组成。为了有效地进行空间结构的损伤检测，系统识别方法是一种很有前途的方法。然而，人们普遍认为系统识别对全局损伤检测有较好的效果，但对局部损伤检测效果不佳。为了克服这一问题，本研究在全局损伤检测和局部损伤检测之间引入了中间损伤检测。在空间桁架的损伤检测中，利用损伤构件定位前动能分布的变化来推断损伤区域。在空间框架的损伤检测中，先对受损层进行定位和评估，然后再确定受损层内的受损部分。此外，为了提高损伤检测方案的准确性和鲁棒性，除了激发全局和低模态振动外，还激发了局部和高模态振动。在空间桁架的情况下，进行了一系列冲击锤试验，以便在一系列振动台试验后激发局部振动。对于空间框架，在实验模型中安装了一个小型振动发生器，以便在一系列振动台试验后激发高模态振动。结果表明，采用多级损伤检测方案的系统识别方法对于空间结构的局部损伤检测仍然是非常有用的。

This study is concerned with the damage detection of space truss and frame structures by using small-scaled experimental models. These structures are composed of a large number of members. To carry out the damage detection of space structures effectively and efficiently, a system identification approach is promising. However, it is widely believed that the system identification is good for global damage detection but not so useful for local damage detection. To overcome this problem, in this study, the intermediate damage detection is incorporated between global and local damage detection. In the case of the damage detection of space trusses, the damaged region is inferred by using the change in the kinetic energy distribution before the pinpoint of damaged members. In the case of the damage detection of space frames, the damaged story is located and evaluated before the pinpoint of damaged portions within the damaged story. To enhance the accuracy and robustness of the damage detection scheme, moreover, local and higher-mode vibrations are excited in addition to global and lower-mode vibrations. In the case of space trusses, a series of impact hammer tests have been performed in order to excite local vibrations after a series of shaking table tests. In the case of space frames, a small vibration generator is installed in the experimental model in order to excite higher-mode vibrations after a series of shaking table tests. It is shown that the system identification approach is still very useful for the local damage detection of space structures by using a multi-level damage detection scheme.