论文部分内容阅读
提出了基于迭代学习控制模型的覆盖件模具拉深筋优化算法,极大地提高了优化效率。利用成形状态函数,成形质量函数和学习律函数构建工艺参数优化的迭代控制模型。将该模型应用到拉深筋阻力值优化中,利用有限元模拟代替很难显示表达的状态函数,预测给定工艺参数方案下板料成形后的应力应变状态。根据单元的应变状态,定义拉深筋线段的局部缺陷程度为成形质量函数,评价拉深筋周围的成形质量好坏。学习律函数不仅参考拉深筋段周围的成形质量偏差确定拉深筋阻力值的改变量,同时还能智能更新学习增益修正拉深筋阻力值的改变幅度,加快了优化收敛速度。通过门内板的算例,证明了该拉深筋优化算法的快速性和实用性。
An optimization algorithm based on the iterative learning control model for the drawbead of the die was put forward, which greatly improved the optimization efficiency. By using the forming state function, the forming mass function and the learning law function, an iterative control model is established to optimize the process parameters. The model is applied to the optimization of resistance value of drawbead. The finite element simulation is used to replace the state function which is difficult to display to predict the stress-strain state of sheet metal under given process parameters. According to the strain state of the element, the degree of local defects in the segment of drawbead is defined as the forming quality function, and the forming quality around drawbead is evaluated. The learning law function not only determines the change of resistance value of drawbead with reference to the deviation of the forming quality around the drawbead, but also intelligently updates the learning gain to correct the changing range of drawbead resistance and accelerates the optimization convergence speed. Through the example of the door inner panel, the fastness and practicability of the drawbead optimization algorithm are proved.