基于多宽度高斯核函数模型的十字形件冲压成形工艺参数优化

Research on parameter optimization of cross-part deep-drawing based on Gaussian kernel function with multi-width

基于试验设计的板料冲压成形工艺参数优化局限在有限的参数组合上,利用响应面方法可以进行全局范围的优化设计。针对一般响应面方法在预测精度上的不足,提出多宽度高斯核函数响应面方法。通过正交试验获得各因素的影响程度指数,利用均匀试验获得合适的设计参数样本,建立基于多宽度高斯核函数的冲压成形响应面模型。对十字形件的冲压成形进行工艺参数优化,结合遗传算法求解满足质量要求的最优工艺条件,验证表明,该方法对提高板料成形质量有效。

Parameter optimization of sheet metal forming based on experiment design is confined to limited parameter combination,while using response surface method can achieve the parameter optimization in globe range.Due to insufficient prediction accuracy in general response surface method,a response surface method was suggested based on Gaussian kernel function with multi-width.Influence degree index was acquired by orthogonal design,and rational experimental samples were acquired by uniform design.The response surface for sheet metal forming was built up based on Gaussian kernel function with multi-width.Parameter optimization of cross-part deep-drawing was performed.By integrating with genetic algorithm,a process with satisfied quality was found,which verified the effectiveness of the method.