基于GS理论与GA-BP网络的板料多目标成形工艺优化

Optimization of Forming Process for Sheet Metal Multi-objective Based on GS Theory and GA-BP Network

成形过程参数在板料冲压成形过程中具有较大的波动性,这种波动可能使得零件出现质量缺陷。本文提出了基于GS理论与GA-BP网络的板料多目标成形稳健优化设计方法。以汽车后纵梁为实例,确定影响其成形质量的7个因素,建立三个目标评价函数,通过正交试验设计和有限元分析得到关联度数据样本,对关联度进行极差分析得出最主要影响因素为压边力。然后将试验数据通过GA-BP网络建立了BHF-μ响应面,并获得该零件冲压时的压边力最优取值范围为350～450kN及最佳成形过程参数组合。通过实际生产验证,该方法对提高板料冲压成形过程的可靠性是有效的。

In sheet metal forming process, with being much greater volatility for the forming processing parameters, it is possible to arise the quality defects of parts. At present a robust optimization design method for sheet metal multi-objective forming was suggested based on the GS theory and GA-BP network. The automotive back rail was selected as an example, and seven factors affecting the forming quality and three goal evaluation functions were determined. The correlations of sample data were obtained by orthogonal test design combined with the FEM. Finally, the main influence factor BHF is found by range analysis of the correlations. The response surface composed of the BHF-μ from 350 kN to 450 kN is established by GA-BP network which is constructed by test data. At the same time, the optimal range of the BFH and the best combination of metal forming process parameters are selected. Applications prove that the method is effective for reliability design of sheet metal forming process.