The Research Of Frmae Structure Damage Detection Based On Lifting Wavelet Transform

Along with the development of economy, civil engineering structure trends to be more large and complicated. The health status of the structure is directly related to the people's lives and property security, therefore the real-time detection of structural damage has become a civil engineering research hotspot. Lifting wavelet transform is the continuation and development of traditional wavelet theory. As a kind of excellent time-frequency analysis tools, lifting wavelet transform, extended from theoretical research to practical application, focuses mainly on signal analysis, image processing, the mechanical fault diagnosis, etc, but the application in civil engineering are less. So it has very important practical significance if lifting wavelet transform can be used to identify the damage features of civil engineering structure.First of all, this paper proposes the basic problems of damage identification in civil engineering structure, summarizes some existing structural damage identification methods. Above all, it mainly elaborates the recognition method based on the wavelet analysis, and introduces its research situation at home and abroad.Secondly, it further introduces the basic theories of lifting wavelet, and enumerates the basic steps of lifting algorithm, which is reflected by multiple phase matrix in the later illustration.Then, this paper gives a damage index by details signal of lifting wavelet, and make use of this index to identify the damage in the rectangular cantilever beam and framework structure. The results showed that details signal index, obtained from ascension transform of the original acceleration signal, can accurately identify the time of damage, but can not improve the location and degree of damage.Finally, in order to identify damage location and damage degree of the cantilever beam model and the framework structure, this paper selects lifting wavelet energy sensitive to structural damage as observation index, through carrying on the different types of Mahalanobis distance calculations using lifting wavelet energy, then establishes damage location index and damage degree index based on distance of Mahalanobis. The result not only effectively identify the two kinds of structures' damage location and damage degree by distance of Mahalanobis distance index, but also show that the recognition method using Mahalanobis distance to test the structural damage is feasible.