车辆悬架纵臂冲压工艺的改进花授粉算法优化  

Optimization on stamping process of vehicle suspension longitudinal arm based on improved flower pollination algorithm

作　　者：刘宝生 邓三鹏[2] Liu Baosheng;Deng Sanpeng(Mechanism Engineering Institute,Tianjin Transportation Technical College,Tianjin 300110,China;Mechanism Engineering Institute,Tianjin University of Technology and Education,Tianjin 300222,China)

机构地区：[1]天津交通职业学院机械工程学院,天津300110 [2]天津职业技术师范大学机械工程学院,天津300222

出　　处：《锻压技术》2021年第11期130-136,共7页Forging & Stamping Technology

基　　金：国家青年自然科学基金资助项目(61301040)。

摘　　要：为了减小车辆悬架纵臂的冲压减薄率、提高产品的加工合格率,提出了冲压工艺参数的改进花授粉算法优化方法。介绍了悬架纵臂三维模型和冲压成形过程。使用有限元仿真分析了开裂临界状态下的等效应力分布和厚度分布,并确定了开裂危险位置。以减小危险位置减薄率为目标建立了优化模型,选择了上模压力和摩擦因数作为优化参数。使用最优拉丁超立方抽样设计了80组实验,建立了2输入1输出的BP神经网络结构,用于拟合工艺参数与质量参数的非线性关系。对花授粉算法的全局搜索方法进行了改进,提高了算法的搜索能力,利用改进花授粉算法求解优化模型得到了最佳参数组合。经验证,优化后悬架纵臂冲压件危险位置的厚度均值为1.637 mm、标准差为0.091 mm;产品合格率由现行的60%提高到了97%。In order to reduce the thinning rate of vehicle suspension longitudinal arm in the stamping and improve the processing qualification rate of products, the optimizated method of stamping process parameters based on flower pollination algorithm was proposed, and the 3D model and the stamping process of suspension longitudinal arm were introduced. Then, the distributions of equivalent stress and thickness in the critical state of cracking were analyzed by finite element simulation, and the critical location of cracking was determined. Furthermore, the optimization model was established with the goal of reducing the thinning rate of dangerous locations, the upper die pressure and friction coefficient were chosen as the optimization parameters. Optimal Latin hypercube sampling was used to design eighty groups of experiment, and BP neutral network structure with two inputs and one output was established to fit the nonlinear relationship between process parameters and quality parameters. Finally, the global search method of flower pollination algorithm was improved to enhance the search ability of the algorithm, and the optimal parameters combination was obtained by using the improved flower pollination algorithm to solve the optimization model. It is clarified that after optimization, the average thickness of staming parts for suspension longitudinal arm at the dangerous location is 1.637 mm, the standard deviation is 0.091 mm, and the product qualification rate increases from the current 60% to 97%.

关 键 词：悬架纵臂 冲压成形 开裂临界状态 改进花授粉算法 减薄率 

分 类 号：TG386[金属学及工艺—金属压力加工]