摘要
针对传统群智能优化算法求解柔性作业车间调度问题时易陷入局部最优且寻优能力不足的困境,以最小化最大完工时间为目标提出一种离散的食肉植物算法。首先,为提高初始种群的多样性,提出了3种初始化种群策略;其次,为提高算法各时期的搜索能力,为生长因子设计了一种自适应策略,并对植物执行了交叉以及基于4种邻域结构的贪婪变异操作;最后,通过对Brandimarte基准问题进行仿真并与其他文献算法进行对比,证明了所提算法在收敛速度和求解质量方面都具有较好的性能。
Aiming at the shortcomings of traditional swarm intelligent optimization algorithms for solving flexible job shop scheduling problems such as being prone to falling into local optimizations and insufficient optimiza⁃tion capabilities,a discrete carnivorous plant algorithm is proposed with the goal of minimizing the maximum completion time.Firstly,three initial population strategies are proposed in order to improve the diversity of the initial population.Secondly,an adaptive strategy for growth factors was designed in order to improve the search ability of the algorithm at each stage,and crossover and greedy mutation operations based on four neighborhood structures were performed on plants.Finally,the Brandimarte benchmark problem is simulated and compared with other literature algorithms.It is proved that the proposed algorithm has good performance in terms of con⁃vergence speed and solution quality.
作者
宋存利
李金泰
SONG Cunli;LI Jintai(School of Software,Dalian Jiaotong University,Dalian 116052,China;School of Computer and Communication Engineer-ing,Dalian Jiaotong University,Dalian 116028,China)
出处
《大连交通大学学报》
CAS
2024年第4期113-120,共8页
Journal of Dalian Jiaotong University
基金
辽宁省教育厅科学研究计划项目(LJKZ0489)。
关键词
柔性作业车间调度
最小化最大完工时间
食肉植物算法
自适应生长因子
混合算法
flexible job shop scheduling
minimize maximum completion time
carnivorous plant algorithm
a⁃daptive growth factor
hybrid algorithm
