基于粒子群遗传算法的航天产品装配顺序优化方法  

Assembly Sequence Planning for Aerospace Products Based on Particle Swarm Optimization and Genetic Algorithm

作　　者：张丹[1] 左敦稳[1] 焦光明[1,2] 薛善良[1] 李建平 

机构地区：[1]南京航空航天大学机电学院,江苏南京210016 [2]南京晨光集团有限责任公司,江苏南京210006

出　　处：《兵工学报》2010年第9期1228-1234,共7页Acta Armamentarii

基　　金：国防"十一.五"预研资助项目

摘　　要：航天产品的装配顺序优化(ASPFAP)具有多目标和非线性的特点,针对传统算法在该问题求解上的不足,将粒子群算法和遗传算法结合起来(PSO-GA),提出一种新的面向航天产品的装配顺序优化方法。使用优先约束关联模型(APCRM)来描述零件间的优先约束关系和关联关系;研究了粒子群遗传算法的基因组、染色体以及粒子的编码表达方法;综合考虑装配连续性、装配资源和仪器设备的影响,提出了有工程意义的适应度函数的表达式;根据APCRM生成随机的可行初始装配序列,并利用粒子群算法重构遗传算法的交叉算子对装配顺序进行优化。实例表明该方法有较好的收敛性和稳定性,优化结果具有实际工程意义。In order to solve the assembly sequence planning for aerospace products（ASPFAP）,which is multi-objective,non-linear and difficult to be solved by the traditional algorithms of ASP,a new method was presented based on particle swarm optimization and genetic algorithm（PSO-GA）.The assembly precedence constraint relationship model（APCRM）was studied;the code representations of genomes,chromosomes and particles were studied;the fitness function with engineering significance was presented by comprehensive consideration of assembly continuity,assembly resource and influence of instrument and equipment;the geometric feasible assembly sequences were initialized according to the APCRM and optimized based on PSO-GA in which the GA＇s crossover operator was reconstructed by PSO.An application case was studied to demonstrate good convergence,stability and actual engineering significance of the proposed algorithm.

关 键 词：机械制造工艺与设备 装配顺序优化 粒子群遗传算法 优先约束关联模型 交叉算子 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]