基于改进花授粉算法的航空发动机装配总体规划
Overall planning of aero-engine assembly based on improved flower pollination algorithm
章斌 1卢洪义 1宋汉强 2刘舜 1杨禹成 1桑豆豆1
作者信息
- 1. 南昌航空大学飞行器工程学院,南昌 330063
- 2. 海军研究院上海研究所,上海 200000
- 折叠
摘要
针对航空发动机结构复杂、零件数量多且装配效率低、装配成本高的问题,提出了一种改进花授粉算法(improved flower pollination algorithm,IFPA)的装配顺序优化方法.以装配优先性、装配稳定性、装配聚合性、装配重定向性和基础部件位置为影响因子构建优化目标评价体系,采用了不同的表示方案、反对立学习的初始种群生成、动态调整的转换概率,在全局授粉和局部授粉规则中引入了均匀变异和精英变异,并加入遗传突变.运用在航空发动机低压压气机装配规划上,验证了 IFPA的有效性,并讨论了 IFPA的参数影响,并同粒子群算法、遗传算法、蚁群算法和花授粉算法进行比较,该算法找到最优序列的概率分别提高了41％、42％、41％和20％,验证了 IFPA在求解装配序列规划问题上的优越性.
Abstract
In view of the problems of complex structure,large number of parts,low assembly efficiency and high assembly cost of aero-engine,an assembly sequence optimization method based on improved flower pollination algorithm(IFPA)was proposed.The optimization target evaluation system was constructed with the influence factors of assembly priority,assembly stability,assembly aggregation,assembly redirection and basic component position.Different representation schemes,initial population generation against independent learning,and dynamically adjusted transition probability were adopted,uniform and elite variation was introduced in global and local pollination rules,and genetic mutation was added.The effectiveness of IFPA was verified by applying it to the assembly planning of aero-engine low-pressure compressor,and the parameter influence of IFPA was discussed.And compared with particle swarm algorithm,genetic algorithm,ant colony algorithm and flower pollination algorithm,the probability of finding the optimal sequence increased by 41％,42％,41％and 20％,respectively,which verified that IFPA can solve the assembly sequence planning superiority in question.
关键词
航空发动机压气机/装配序列规划/花授粉算法/均匀变异/精英变异Key words
aero-engine compressor/assembly sequence planning/flower pollination algorithm/uniform mutation/elite mutation引用本文复制引用
基金项目
江西省自然科学基金(20201BBE51002)
江西省研究生创新专项(YC2021-S685)
出版年
2024
NETL
NSTL
万方数据