模糊神经网络与启发式算法相结合的制造单元构建方法  

Manufacturing cell formation approach combing fuzzy neural network with heuristic algorithm

作　　者：王国新[1] 阎艳[1] 宁汝新[1] 王爱民[1] 

机构地区：[1]北京理工大学机械与车辆工程学院,北京100081

出　　处：《计算机集成制造系统》2008年第11期2175-2183,共9页Computer Integrated Manufacturing Systems

基　　金：国防基础科研资助项目(B0920060901)~~

摘　　要：为提高制造单元构建效率和制造单元的可实施性,提出了模糊自适应谐振神经网络与启发式算法相结合的制造单元构建方法。该方法以加工相似模式匹配思想为指导,将制造单元的构建分为两个阶段。首先对标准模糊自适应谐振神经网络进行改进处理,并以此构建初始的工件族和设备集合;然后,在初始工件族和设备集合的基础上,建立基于规则的启发式资源优化分配算法,解决制造单元的设备共享、可选设备分配和设备负荷不均等问题。最后,以某企业机加车间生产的典型工件为例,验证了算法的有效性。To improve the efficiency of manufacturing cell formation and its practicability, a manufacturing cell formation ap proach combing Fuzzy Adaptive Resonance Theory Neural Network （FARTNN） and heuristic algorithm was proposed. Guided by ideas of matching of similar machining modes, this approach was divided into two stages. Firstly, the improved FARTNN was used to form the basic part family and equipment sets. Then, based on the basic part family and equipment sets, a rule-based heuristic algorithm for resource allocation was presented to resolve the problems of equipment sharing, alternative equipment allocation and unbalanced equipment load in the manufacturing cell, etc. Finally, the validity of the approach was proved by its application in a workshop of a manufacturing enterprise.

关 键 词：模糊神经网络 启发式算法 可重构制造系统 制造单元 

分 类 号：TH18[机械工程—机械制造及自动化] TP391[自动化与计算机技术—计算机应用技术]