Application of a Gaussian filter to smooth optimization of free form grid shells скачать в хорошем качестве

Application of a Gaussian filter to smooth optimization of free form grid shells

Parallel Session 53, Optimisation methods for analysis and design of roof structures (WG 13) Hiroki Tamai (Nikken Sekkei) presents his work on optimal geometry. In design practice, structurally optimal geometry is required to be as smooth as it is initially created by the architect. It is usually the case for the free form of a grid shell structure. This video presents an application of Gaussian filters inside the optimization process of structural geometry. Two-dimensional convolution of a roof geometry and Gaussian kernel effectively spread the variation of a nodal coordinate in the design variables to the adjacent nodes by the effect of weighted average. Thus, no node can be modified independently. It is performed simultaneously whenever optimization algorithm updates the geometry. In addition to maintaining a pleasant smoothness in the grid shell, the low pass filtering effects, as a consequence, can reduce the chance that the solution is trapped in local minima. Alternatively, you can read the paper by accessing our conference proceedings. 📚 Use the link https://doi.org/10.15126/900337 to download your digital copy and let yourself be #inspired! Watch his presentation on "Application of a Gaussian filter to smooth optimization of free form grid shells" from the Spatial Structures 2020/21 Conference. #IASS #IASS2020_21 #SpatialStructures #Architecture #Structures #Engineering #LightweightStructures #CivilEngineering #StructuralEngineering #Conference #Construction #computationaldesign #nextgeneration #parallelsessions #optimisation #methods #analysis #design #roof #roofstructures #workinggroup #gridshells