Finite Element Mesh Model Optimization By Sample Learning

Finite element analysis has been widely used in various fields of industrial product analysis. Of all phases of FEA, mesh model generation is the key ingredient, which directly influences the accuracy and speed of FEA. In order to achieve high quality mesh, a lot of topology-based mesh optimization methods have been applied. However, they are all quite time consuming. Therefore, this paper brings artificial intelligence into mesh optimization in order to improve efficiency. The contributions of this paper are as follows:(1)This paper presents a SVM-based mesh optimization method. We design the framework for such method based on the analysis of mesh optimization and SVM.(2)This paper presents a SVM-based flip, which derives from the general method proposed above. We firstly find features for 2-2 flip,2-3 flip and 3-2 flip respectively, following by using these features to build SVM model. Finally, the method is verified and analyzed by some experiments.(3)This paper presents a SVM-based edge collapse, which derives from the general method proposed above. We firstly find feature for edge collapse, then use these features to build SVM model. Finally, the method is verified and analyzed by some experiments.The experiment result shows the SVM-based mesh optimization method can improve the consuming time without losing mesh quality.