首页 | 官方网站   微博 | 高级检索  
     

基于遗传算法和神经网络的冷挤压工艺参数模糊优化设计
引用本文:石峰,娄臻亮,张永清,陆金桂. 基于遗传算法和神经网络的冷挤压工艺参数模糊优化设计[J]. 机械工程学报, 2002, 38(8): 45-49
作者姓名:石峰  娄臻亮  张永清  陆金桂
作者单位:1. 上海交通大学模具CAD国家工程研究中心,上海,200030
2. 南京化工大学
基金项目:教育部优秀青年教师基金,国家自然科学基金(60175019),上海市高等学校青年科学基金资助项目
摘    要:将模糊优化思想引入冷挤压工艺参数优化设计,以冷挤压模具的四个关键工艺参数为设计变量,以最终挤压件的各处材料的损伤值为目标函数,建立了冷挤压工艺参数模糊优化模型。提出了利用神经网络进行材料损伤近似计算的策略,从而形成了基于遗传算法和神经网络的冷挤压工艺参数模糊优化方法。对冷挤压工艺参数模糊优化模型进行了求解,优化结果表明模糊优化思想能提高冷挤压工艺设计质量。

关 键 词:遗传算法  模糊优化  神经网络  冷挤压工艺参数优化  正交试验
修稿时间:2001-04-16

FUZZY OPTIMIZATION OF COLD EXTRUSION MOLD BASED ON GENETIC ALGORITHM AND NEURAL NETWORK
Shi Feng Lou Zhenliang Zhang Yongqing. FUZZY OPTIMIZATION OF COLD EXTRUSION MOLD BASED ON GENETIC ALGORITHM AND NEURAL NETWORK[J]. Chinese Journal of Mechanical Engineering, 2002, 38(8): 45-49
Authors:Shi Feng Lou Zhenliang Zhang Yongqing
Affiliation:Shi Feng Lou Zhenliang Zhang Yongqing (Shanghai Jiaotong University)Lu Jingui (Nanjing University of Chemical Technology)
Abstract:The idea of fuzzy optimization is introduced in the optimization of cold extrusion mold process parameters. An approach of fuzzy optimization of cold extrusion mold process parameters based on genetic algorithm and multiplayer neural network is presented. Firstly, the fuzzy optimization model of cold extrusion mold is introduced. Secondly, the fuzzy optimization strategy using genetic algorithm and multiplayer neural network is discussed, including constructing an approximate model with artificial neural networks and solving the fuzzy optimization model using genetic algorithms. In constructing fitness function, the weight factors are determined by orthogonal numerical experiments. Finally, the optimization results demonstrate the fuzzy optimization method can improve the quality of the cold extrusion mold.
Keywords:Genetic algorithm Fuzzy optimization Neural network Optimization of cold extrusion mold process pa-rameters Orthogonal numerical experiments
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号