Dynamic Responses Analysis Of Structures Using Wavelet Transforms

There are a lot of structures subjected to not only static loads, but also dynamic excitation in real work and lives. And in many case, destructions under dynamic loads are worse by far than that under static loads. For such kind of structures and systems, in order to improve their dynamic characteristics, with intend to control vibration or to ensure they can work on the safe side in dynamic condition, dynamic design must be applied. Different from the problem on static intensity, the relation between structural dynamic response and design variables is even more complex, and is high nonlinear. The analytical computational work for structural dynamic response, especially for the transient response, is so much that computational efficiency and computational precision are both problems that deserve to be research carefully.Dynamic analysis of the structures against the original earthquake record is, in general time consuming,which affect real time disposal for data to a great extent. It has practical significance that searching out a arithmetic shorten time for disposal. If expanding time interval (namely reducing sampling) in a certain extent, calculating amount of step and the computational work can be reduced, the purpose to reduce the execution time may be achieve.The low frequency content is the most important part of signal, because it contains most of the energy of the signal and the characteristics of the low frequency content are similar to those of the original signal. Therefore the low frequency of original record can only be used for dynamic analysis. Wavelet transform can decompose accelerogram ground acceleration into two parts including low frequency and high frequency. Because of downsampling in course of wavelet transforms signal, each part of decomposition of the signal geminate to reduce, it just accord with the idea of increasing time interval.In this paper, it is a attempt that the dynamic analysis of structures is carried out against this low frequency part with lesser sampling points of wavelet decomposition, then data, and the results are compared with dynamic computation using the original earthquake record and result to multiple time-step length directly. The method of data processing that may cut down computational burden based on demand, and which precision should be up to the mustard, is tried to find.To reduce the computational work, fast wavelet transform is used. To achieve...