A Knowledge-based Mesh Generation System for Forging Simulation
Osamu Takata, Koukichi Nakanishi, Nariaki Horinouchi, Hiroshi Yano, Tadao Akashi, Toyohide Watanabe · Industrial and Engineering Applications of Artificial Intelligence and Expert Systems · 2022
We have developed a knowledge-based system GENMAI (Artificial Intelligence Mesh GENerator) for automatically generating two-dimensional meshes, which is easily applicable to various fields of computational dynamics with structured meshes. We analyzed and modeled experts’ mesh generation procedures in the simulation processes of forging deformations, vehicle aerodynamics and engine in-cylinder flows. This paper describes the outline of GENMAI focusing on the implementation techniques of search and inference, and then shows the functional ability of GENMAI applying to the forging simulation. The characteristics of GENMAI are as follows. (1) Plural solutions can be efficiently obtained at the same time in the search phase, using the global dependency and local dependency. (2) The meta-level inference method and its knowledge representation method are very applicable to various fields of analyses. We have applied GENMAI to forging simulation, and found it to have a prospect for practical use.