The Application Of Fuzzy Clustering Algorithm In The Defects Prediction And Optimization Of Stamping Process

Sheet metal forming technology possesses many advantages with good dimensional accuracy, interchangeability, high dimensional stability of parts, high production efficiency, saving material and so on. With the rapid development of the automobile, aerospace, shipbuilding and other industries, sheet metal forming technology is used widely more and more. However, sheet metal forming is a complex elastic-plastic, large deformation process with a high degree of geometric nonlinearity, material nonlinearity, non-linear characteristics of the border and so on, therefore how to find the law of sheet metal forming has a very important theoretical and economic value.Fuzzy clustering analysis method which made the clustering accurate and scientific, can extract the potential characteristics of stamping process, provide a new research method for the stamping precess design, has made the detects prediction and optimization of stamping process more accurate and scientific as well as possible.The major directions of this article are the application of Fuzzy clustering algorithm in the defects prediction and optimization of stamping process. A preliminary study and exploration has been done on the application of fuzzy clustering algorithm in the stamping process design and the specific tasks are as follows:Combined with the fuzzy clustering method based on equivalence relations, a practical model for process design was established.Combined the fuzzy clustering algorithm based on objective function with a statistical theory, the establishment of prediction model for V-bending process springback, its VC programming, and the forecast results analysis of V-bending springback have been achieved.The main objective functions of the flanging process were established and the results of finite element simulation was extracted for the calculation of the objective functions. Combined with multi-objective fuzzy planning theory, flanging process parameters are optimized and the results were verified by experiments.