抄録
This paper presents a new inverse analysis method based on wavelet transform theory. For solving inverse problems, it is important to use a priori information efficiently. This method enables to use the local frequency-characteristic of estimates as a priori information which has not be used in conventional methods. Firstly, the estimates are divided into some components by using wavelet transform algorithm. Each component has different frequency-characteristic/resolution. Secondly, according to a priori information, elements in the component are manipulated. Thirdly, those components are identified from low to high frequency-characteristic/resolution component sequentially. Finally, the estimates are reconstructed from those identified components by using inverse wavelet transform algorithm. A numerical simulation is performed to demonstrate the validity and usefulness of the proposed technique.
