融合随机重启爬山算子的改进遗传算法求解FJSP  

Improved genetic algorithm with integrated random-restart hill climbing operator for solving FJSP

作　　者：陈亚铭 潘大志[1,2] Chen Yaming;Pan Dazhi(School of Mathematics and Information,China West Normal University,Nanchong 637009,China;Institute of Computing Method and Application Software,China West Normal University,Nanchong 637009,China)

机构地区：[1]西华师范大学数学与信息学院,南充637009 [2]西华师范大学计算方法与应用研究所,南充637009

出　　处：《现代计算机》2024年第11期29-34,共6页Modern Computer

基　　金：国家自然科学基金资助项目(11871059);四川省教育厅自然科学基金资助项目(18ZA0469);西华师范大学英才科研基金资助项目(17YC385)。

摘　　要：针对传统遗传算法在求解柔性车间调度问题时,存在种群的动态适应能力差、容易陷入局部最优等问题,提出一种融合随机重启爬山算子的改进遗传算法。通过双种群交叉,增强种群间的信息交流能力。引入收敛准则,在维护种群多样性的同时防止种群的优良个体被过度破坏。结合随机重启爬山法的思想进行局部搜索,提升了算法的局部搜索能力。仿真实验表明,所提出的算法在不同规模的问题中,都有着明显的寻优能力。In response to the challenges faced by traditional genetic algorithms when addressing the Flexible Job Shop Scheduling Problem(FJSP),such as poor dynamic adaptability of populations and susceptibility to local optima,proposing an improved genetic algorithm that integrates a random-restart hill climbing operator.Enhancing the ability to exchange information between populations through dual population crossing.Introducing convergence criteria to maintain population diversity while preventing excessive disruption of superior individuals within the population.By incorporating the principles of random-restart hill climbing,the algorithm’s local search capabilities are significantly improved.Simulation experiments demonstrate that the proposed algorithm consistently exhibits strong optimization performance across problems of varying scales.

关 键 词：柔性车间调度 改进遗传算法 接受准则 随机重启爬山算子 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]