连杆衬套强力旋压成形的多目标优化与决策  

Multi-object optimization and decision on power spinning for connecting rod bushing

作　　者：张冰[1] 徐嘉锋 Zhang Bing;Xu Jiafeng(College of Information Engineering,Xuchang Vocational Technical College,Xuchang 461000,China)

机构地区：[1]许昌职业技术学院信息工程学院,河南许昌461000

出　　处：《锻压技术》2022年第2期119-125,共7页Forging & Stamping Technology

基　　金：河南省科技攻关项目(1821022100508)。

摘　　要：为了减小连杆衬套强力旋压成形的圆柱度误差和壁厚偏差,提出了基于加权相对距离的最优解决策策略。介绍了连杆衬套的强力旋压成形工艺,以减小圆柱度误差和壁厚偏差为目标建立了多目标优化模型。使用RBF神经网络和有限元仿真软件拟合了工艺参数与目标参数之间的函数关系,经验证RBF神经网络的拟合精度较高。使用非支配排序遗传算法对多目标优化模型进行求解,得到了Pareto前沿解集。建立了理想解和不理想解的概念,提出了基于加权相对距离的最优解决策方法。经生产验证,优化后的圆柱度误差均值比厂家生产的圆柱度误差均值减小了23.16%,壁厚偏差均值减小了23.94%,且两个质量参数的标准差也有所下降。实验结果说明:优化后的连杆衬套圆柱度误差和壁厚偏差均有所减小,且生产稳定性有所提高。In order to reduce the cylindricity error and wall thickness deviation of connecting rod bushing in power spinning, an optimal solution decision-making strategy based on weighted relative distance was proposed. Then, the power spinning process of connecting rod bushing was introduced, and a multi-objective optimization model was established to reduce the cylindricity error and wall thickness deviation. Furthermore, the functional relationship between process parameters and target parameters was fit by RBF neural network and finite element simulation software, and the fitting accuracy of RBF neural network was verified to be higher. In addition, the multi-objective optimization model was solved by the non-dominant sorting genetic algorithm to obtain Pareto frontier solution set, and the concepts of ideal solution and non-ideal solution were established to propose the optimal solution decision-making method based on weighted relative distance. Through production verification, after optimization the mean value of cylindricity error was reduced by 23.16% compared with the mean value of cylindricity error produced by the manufacturer, the mean value of wall thickness deviation was reduced by 23.94%, and the standard deviation of the two quality parameters was also reduced. The experimental results show that the cylindricity error and wall thickness deviation of the optimized connecting rod bushing are reduced, and the production stability is improved.

关 键 词：连杆衬套 强力旋压 圆柱度误差 壁厚偏差 Pareto前沿解 加权相对距离 非支配排序遗传算法 

分 类 号：TG306[金属学及工艺—金属压力加工]