Damage Identification Research Based On The Improved Cross-model Cross-mode Method

With the development of science and technology, civil engineering structure is becoming more and more complex and huge. When the structure is completed, it is exposed to the external environment and under the action of various complex loads, structure performance will becoming worsen gradually, which is no good for the structure to resist the bad environment and various load. Structural health monitoring could monitor structure's overall performance and stiffness change of local components. Install with the health monitoring system, calamity cased by possible collapse or failure of some component may be avoided.Structural damage identification is the core and an important part of structural health monitoring. In recent years, structural damage identification methods have got a lot of development; all kinds of new methods emerge in endlessly. Methods based on structure frequency were put forward first, then is the identification methods using structure modal, now are the presentence of cross model modal method and neural network, etc. The basic theory of structure damage identification is:the modal parameters of structural are functions of the structure's physical characteristics; the physical characteristics change will cause corresponding vibration response change. Through the analysis of structure vibration response data we can obtain modal parameters change information. This paper devoted to obtain better damage identification methods, in order to apply the health evaluation of the structure and avoid calamity and reduce losses from building collapse.Based on the cross model cross modal method, this paper mainly conducted a study on the several aspects as follow:(1) Started with LANL laboratory's eight degree of freedom system, a structure identification method combining the cross model cross modal method and the sensitivity correction method was put forward. According to the actual existence error between theoretical model and experimental structure, this paper first use experimental modal information to fix a theory model and then applied the cross model cross modal method on damage identification, finally we obtained an ideal effect.(2) A kind of optimization on the cross model cross modal method was put forward. The cross model cross modal method is modified to improve the computational efficiency and economize the storage space. Program realization of the cross model cross modal method is carried out in this article. The method is proven effectiveness. (3) The application of the cross model cross modal method in space frame structure is studied in this article. Cross model cross modal method is a relatively new damage identification method. the author first put forward the method for the purpose of structure model modification. In view of the superiority of this method in damage identification process, this paper explores the damage identification applications in steel space frame structure.The results of the study show that, cross model cross modal method can make full use of the structure's lower order modal to identify damage. In the two conditions of complete and incomplete modal spaces, we can both get perfect damage identification results. This method is physical meaning definitely, facilitate simulation and application. This improved cross model cross modal method put forward in this paper is both effective to identify damage location and degree and effective in saving system resources.