A regularization based method of inversely analyzing the material parameters for the gravity dams is presented in this paper. The measured noise of the displacement often leads to the ill-posed solutions of the dam inversion. A least square scheme was adopted for the inverse analysis by using of the displacement measurements, and the L-curve algorithm was proposed to perform the regularization in the inversion computation. The proposed method actually utilizes the sensitivity of the measured displacements with respect to the sub-regional material parameters. Thus a simplified algorithm was developed to calculate the sensitivity coefficients by using the standard finite element procedure. To investigate the effectiveness of the proposed method, the numerical simulations on a concrete gravity dam were carried out. The results demonstrate that the presented method can accurately obtain the material parameters of the dam by using of the inverse analysis, which has the potential for assessing the integrity of the gravity dams.

In this project, we propose a parallelized DSW algorithm, useful to periodically rebalance an arbitrary unbalanced BST. The proposed algorithm splits the tree into several subtrees and concurrently do ``Tree-to-Vine'' to these subtrees using multiple threads, and finally connects them into a complete vine; It also splits the vine into several semi-equal sized subvines and do ``Vine-to-Tree'' to these subvines using multiple threads, and combines them into a complete balanced tree. Our extensive experiments including BST generation, ``Tree-to-Vine'', and ``Vine-to-Tree'' parallelization comparisons show that this parallelized DSW algorithm performs much better compared to its counterpart sequential DSW algorithm.