基于RBFNN-GA的纵封热封过程工艺参数决策方法研究  

Research on the Decision-Making Method of Process Parameter in Longitudinal and Heat Sealing based on RBFNN-GA

作　　者：彭来湖[1,2] 徐钦栋 PENG Laihu;XU Qindong(Zhejiang Sci-Tech University,Hangzhou 310000,China;Zhejiang Sci-Tech University Longgang Research Institute,Wenzhou 325000,China)

机构地区：[1]浙江理工大学,浙江杭州310000 [2]浙江理工大学龙港研究院,浙江温州325000

出　　处：《软件工程》2022年第10期32-36,共5页Software Engineering

摘　　要：为了提高枕式包装机纵封热封流程生产效率,降低人力成本,提出了一种基于径向基函数神经网络-遗传算法(Radial Basis Function Neural Network-Genetic Algorithm,RBFNN-GA)的工艺参数决策方法。根据热封材料断裂屈服强度确定加工目标热封强度;以热封距离、热封温度、热封速度为工艺参数变量,通过生产样本对RBF神经网络进行训练建立模型,并以此模型为基础,设计遗传算法寻优过程当中的适应度函数;通过遗传算法进行工艺参数的最优求解,并将结果输入到枕式包装机进行加工测试与热封强度的测定。结果表明:根据所述方法得到的测试结果满足实际加工需求,为枕式包装机的智能化升级提供了理论依据。In order to improve the production efficiency of the longitudinal and heat sealing process of the pillow packaging machine,and reduce the labor costs,this paper proposes a process parameter decision-making method based on Radial Basis Function Neural Network-Genetic Algorithm (RBFNN-GA).Firstly,the heat sealing strength of the processing target is determined based on the breaking yield strength of the heat sealing material.Then,taking heat sealing distance,heat sealing temperature and heat sealing speed as process parameter variables,the RBF neural network is trained through production samples to establish a model,based on which the fitness function in the optimization process of genetic algorithm is designed.Lastly,the optimization of process parameters is solved by genetic algorithm,and the results are input into the pillow packaging machine for processing test and heat sealing strength determination.The results show that the test results obtained by the proposed method meet the actual processing needs,and provide a theoretical basis for the intelligent upgrading of pillow packaging machine.

关 键 词：枕式包装机纵封 热封强度 参数寻优 RBF神经网络 遗传算法 

分 类 号：TP241[自动化与计算机技术—检测技术与自动化装置]