静态知识化制造环境下航空发动机装配车间周期性自进化研究  

Periodic Self-evolution of an Aircraft Engine Assembly Workshop under Static Knowledgeable Manufacturing Environment

作　　者：姜天华[1] 严洪森[1] 汪峥[1] 

机构地区：[1]东南大学复杂工程系统测量与控制教育部重点实验室,南京210096

出　　处：《中国机械工程》2015年第7期903-911,共9页China Mechanical Engineering

基　　金：国家自然科学基金资助重点项目(60934008);中央高校基本科研业务费专项资金资助项目(2242014K10031);江苏高校优势学科建设工程资助项目

摘　　要：针对静态知识化制造环境下的航空发动机装配车间,研究了知识化制造系统自进化问题,采用滚动时域方法实现了该装配车间的周期性自进化。根据问题的特点,建立了系统在每个决策时刻的静态子问题的数学模型,提出了基于初始决策方案的滚动规则,进而给出了一种求解自进化问题的两阶段算法。针对所建立的数学模型,设计了一种具有双层结构的遗传算法(Bi GA)。最后通过仿真验证了模型和算法的有效性和可行性,并分析了滚动窗口和滚动步长的大小对生产性能的影响。For an aircraft engine assembly workshop under static knowledgeable manufacturing environment,the self-evolution problem of knowledgeable manufacturing systems was studied.Roll-ing horizon method was adopted to implement the assembly workshop’s periodic self-evolution.On the basis of the characteristics of the problem,a mathematical model of the static sub-problem at each decision moment was established.A rolling rule was proposed based on the initial decision scheme, and a two-phase algorithm was given to solve the self-evolution problem.For the mathematical mod-el,a BiGA was proposed.Finally,simulation results demonstrate that the proposed model and algo-rithms are effective and feasible.The influences of the sizes of rolling window and rolling step on the production performance were also analyzed.

关 键 词：静态知识化制造环境 自进化 滚动时域 双层遗传算法 两阶段算法 

分 类 号：TH165[机械工程—机械制造及自动化]