NEW DEVELOPMENTS IN STRUCTURAL OPTIMIZATION USING ADAPTIVE MESH REFINEMENT AND MULTIPOINT APPROXIMATIONS

摘要: Combining Shape Optimization (SO) with Adaptive Mesh Refinement (AMR) potentially offers a higher accuracy and computational efficiency, especially if the applied target error for AMR is reduced in course of optimization process. The disadvantage that approach rate convergence corresponding processes can be significantly lower as compared to which apply fixed AMR. In present paper so-called Multipoint Approximation Method (MAM) used basis SO conjunction Several techniques improvement rates are presented investigated. Firstly, alternative algorithms determining approximation functions using weighted least squares method focus on weights depend discretization errors. Secondly, different strategies moving resizing search sub-regions space design variables presented. proposed methods il...