一种基于强化学习的作业车间动态调度方法

Production scheduling is critical to manufacturing system. Dispatching rules are usually applied dynamically to schedule the job in a dynamic job-shop. Existing scheduling approaches sel- dom address machine selection in the scheduling process. Composite rules, considering both machine selection and job selection, are proposed in this paper. The dynamic system is trained to enhance its learning and adaptive capability by a reinforcement learning (RL) algorithm. We define the conception of pressure to describe the system feature. Designing a reward function should be guided by the scheduling goal to accurately record the learning progress. Competitive results with the RL-based approach show that it can be used as real-time scheduling technology.     

Job- shop 提前/拖期调度问题的研究

基于模糊控制和遗传算法,提出了求解Ｊｏｂ－ｓｈｏｐ提前／拖期间问题的联合算法,用遗传算法确定可行调度序列,然后用模糊控制器对开工时间加以调整,模糊控制的引入为有效地求解Ｊｏｂ－ｓｈｏｐ提前／拖期调度总理２提供了新的方法,仿真实验证明了联合自救的有效性。     

A Genetic Algorithm Approach to Solving Stochastic Job-shop Scheduling Problems

This paper proposes a method for solving stochastic job-shop scheduling problems based on a genetic algorithm. The genetic algorithm was expanded for stochastic programming. In this expansion, the fitness function is regarded as representing fluctuations that may occur under stochastic circumstances specified by the distribution functions of stochastic variables. In this study, the Roulette strategy is adopted for selecting the optimum solution in terms of the expected value. Within this algorithm, it is expected that the individual that appears most frequently must give the optimum solution. The effectiveness of this approach is confimed by applying it to stochastic job-shop scheduling problems. I compare the approximately optimum solutions found by this approach with the truly or approximately optimum solutions obtained by other conventional methods, and discuss the performance and effectiveness of this approach.     

Job-shop排序问题的遗传强化学习算法

针对Job-shop排序问题的复杂性,提出一种遗传强化学习算法对其求解。通过引入多个随机变量,把Job-shop排序问题转换成多阶段决策问题,通过仿真手段构建作业排序问题模型环境,求取系统性能指标并保证解的可行性。设计一个多智能体Q-Learning算法和遗传算法相结合的算法用于解决Job-shop排序问题。仿真优化实验结果验证了该算法的有效性。     

LR方法基于车间调度问题的面向对象设计

针对拉格朗日松弛方法解决不同车间调度问题时,对问题的依赖性强,算法实现复杂的情况,通过分析拉格朗日方法解决不同车间调度问题的特点,提出了拉格朗日算法面向时象的设计方法,并开发了通用的类模块;面向对象的模块关系和类层次使得算法可扩展性强,便于改进。仿真结果表明,用户可以方便地实现拉格朗日方法对多种车间调度问题的仿真,大大提高了代码的可重用性和软件的通用性。     

基于混合遗传算法的作业车间调度问题的研究与仿真

作业车间调度问题(JSP)是一类典型的NP-hard问题,遗传算法(GA)由于其隐合并行性和全局解空间搜索两大优点而成为解决JSP问题的常用工具.但是,由于JSP问题本身的特点,普通遗传算法难以在解此类问题时得到满意解,最突出的问题就是过早收敛于某一局部最优解,使算法效率降低.在此基础上,将遗传算法和模拟退火算法相结合,提出了一种基于遗传和模拟退火的混合算法,该算法将模拟退火算法赋予搜索过程时变性融入其中,具有明显的概率跳跃性,并选取了典型问题进行分析和仿真研究.仿真结果表明,与传统的遗传算法相比该方法是行之有效的.     

基于成本的车间调度模型及遗传算法求解

制造企业的成本优化，无论在学术界还是工业界，始终是一个热点研究课题。从成本优化思想出发，建立了基于成本的多产品切换作业车间调度问题的经济指标模型，并应用具有新型交叉、变异算子的遗传算法作为求解该调度问题的方法，其目的是使企业的综合生产成本（主要包括多产品切换成本、产品作为在制品的存储成本以及产品拖期惩罚等）最低。仿真实例的运行结果表明了该调度模型和优化算法的有效性，且适用于过程工业及离散工业的生产调度系统，实现生产调度和经济效益的有效集成。     

柔性作业车间调度问题的集成启发式算法

柔性作业车间调度问题,包括路径分配和加工排序2大子问题,是组合优化理论和实际生产管理的重要研究方向。作为传统作业车间调度的扩展,柔性作业车间调度问题的内在复杂性(强NP-Hard)使得传统的最优化方法难以有效求解。文章针对以多目标权重和最优为目标的柔性作业车间调度问题,提出基于过滤定向搜索的集成启发式算法,设计改进了节点分枝策略和局部/全局评价函数,能同时解决2大子问题。通过实例仿真,对算法性能进行比较分析和评价,结果表明了算法的可行性和有效性。     

基于调度规则和免疫算法的作业车间多目标调度

利用动态在线调度方法对动态环境下的作业车间进行研究,采用优先级调度规则对大量调度案例进行求解,针对7个调度目标,从备选调度规则集中选出了单个目标下性能最优的调度规则;为实现调度规则的动态选择以适应多目标调度,基于免疫系统中的独特型网络理论,设计了一种免疫调度算法．根据算法,定义了有效的抗体和抗原结构,并通过抗体间亲和力计算、抗体浓度计算、抗体选择等关键步骤,实现对调度规则的动态控制．仿真测试数据表明,所设计的免疫调度算法能根据不同的车间情况,快速选出不同的调度规则满足多个调度目标,有效解决了作业车间多目标调度问题．     

单件小批量生产排产逆序递推法应用研究

针对单件小批量型生产企业,为了保证交货期和控制原材料、在制品库存,提出一种生产排产算法--逆序递推算法.本文在算法中考虑到了生产扰动对交货期的影响,并结合实例作出逆序递推排产甘特图来指导生产.     

柔性作业车间调度的多Agent协商策略

柔性作业车间调度问题是经典作业车间调度问题的扩展。为此,提出一种新的基于招投标的多Agent协商调度策略,并研究各Agent协商时的价格函数。系统主要由工件Agent和机器Agent组成,工件Agent通过招投标的方式,选择合适的机器完成加工任务,机器Agent按照市场机制通过自由竞争获得工件的加工权,根据基于规则的调度策略处理工件。用Java设计仿真实验程序,并通过实验验证所提价格协商函数的有效性。     

Resource Constraints for Preemptive Job-shop Scheduling

This paper presents an experimental study of constraint propagation algorithms for preemptive scheduling. We propose generalizations of non-preemptive constraint propagation techniques (based on timetables, on disjunctive constraints, and on edge-finding) to preemptive and mixed problems, i.e., problems in which some activities can be interrupted and some cannot. Another approach relies on incremental flow-based techniques. We theoretically compare these approaches and present an experimental comparison based on a branch and bound procedure for the preemptive variant of the job-shop scheduling problem. We show that both edge-finding and flow-based techniques allow the resolution of hard problem instances, including the preemptive variant of the famous FT10.     

遗传算法求解多目标柔性Job-shop问题

本文描述了基于可变机器约束的多目标柔性Job-shop调度问题模型,并应用一种改进的遗传算法进行求解。我们采用了表示工序先后顺序及机器选择的二维编码方式,以多目标优化函数为度量,通过三种遗传操作扩展后代的多样性和算法的搜索空间。仿真结果验证了该算法能有效解决多目标优化问题。     

求解柔性作业车间调度问题的改进蝙蝠算法

针对柔性作业车间调度问题,在研究和分析蝙蝠算法的基础上,提出一种改进蝙蝠算法来求解．为了有效地表达出工序与粒子种群之间的关系,提出一种单层整数编码策略．在粒子的速度和位置方面,算法重新定义速度和位置的相关算子．为了克服基本蝙蝠算法固定参数不足的缺点,重新调整惯性权重的值,提出一种呈指数递减的惯性权重策略．针对具体生产实例进行验证,实验数据表明,该改进算法在求解柔性作业车间调度问题上具有良好的性能,是一种有效的调度算法.     

一种求解单件车间调度问题的单亲遗传算法

针对单件车间凋度问题,设计一种基于整数编码的单亲遗传算法。该算法既具有单亲遗传算法运算量小、不存在“早熟收敛”现象等优点,在编码中又体现了单件车问调度的“保序性”等工艺约束条件,增强了调度算法的整体性能。对不同规模的FT类问题的仿真结果表明,该算法的收敛速度随着问题规模的增大而加快,并在求解单件车问调度问题时能得到稳定的满意解。     

A new Reinforcement Learning-based Memetic Particle Swarm Optimizer

Developing an effective memetic algorithm that integrates the Particle Swarm Optimization (PSO) algorithm and a local search method is a difficult task. The challenging issues include when the local search method should be called, the frequency of calling the local search method, as well as which particle should undergo the local search operations. Motivated by this challenge, we introduce a new Reinforcement Learning-based Memetic Particle Swarm Optimization (RLMPSO) model. Each particle is subject to five operations under the control of the Reinforcement Learning (RL) algorithm, i.e. exploration, convergence, high-jump, low-jump, and fine-tuning. These operations are executed by the particle according to the action generated by the RL algorithm. The proposed RLMPSO model is evaluated using four uni-modal and multi-modal benchmark problems, six composite benchmark problems, five shifted and rotated benchmark problems, as well as two benchmark application problems. The experimental results show that RLMPSO is useful, and it outperforms a number of state-of-the-art PSO-based algorithms.     

工件工序可并行的作业车间调度模型与求解

针对传统作业车间调度模型没有考虑工件工序存在并行性的不足,提出一种以最小化完工时间为目标的工件工序可并行作业车间调度模型,且在模型中考虑了工序加工设备柔性;设计了基于遗传算法的调度算法,其中染色体编码采用分段编码方式,并提出一种适用于工件工序存在并行性的染色体解码方法.实验结果表明,文中算法能够有效地解决工件工序可并行的作业车间调度问题.     

含有运输小车的柔性作业车间调度研究

在实际的柔性作业车间调度中,不但工件需要加工时间,而且工件在各个机器之间利用AGV(自动导引小车)转移也需要占用一定的时间,因此对柔性作业车间调度中考虑AGV运输时间的研究更具有实际意义.针对此问题,本文建立含有AGV的柔性作业车间调度的数学模型,针对问题自身特点对遗传算法进行改进,引入局部搜索策略加强局部寻优能力,将...     

多工艺路线车间作业调度遗传算法实现

针对简单遗传算法在解决作业车间调度问题时只适用于简单问题的局限,研究了多工艺路线的批量调度遗传算法实现,论述了3种提高生产效率的调度策略,即采用最小批量原则对零件进行分批调度生产;将批量准备时间和零件加工时间相分离,在工件到达加工机床前做好批量加工准备;在生产加工过程中,将同批加工零件进行多次机床间转移,缩短后续机床的等待时间.同时将工序优先级调度算法加入到简单遗传算法,提出了一种全局优化的多工艺路线批量生产调度混合遗传算法.仿真结果表明,该调度算法能取得较好的效果.     

A Reinforcement Learning Approach to Interval Constraint Propagation

When solving systems of nonlinear equations with interval constraint methods, it has often been observed that many calls to contracting operators do not participate actively to the reduction of the search space. Attempts to statically select a subset of efficient contracting operators fail to offer reliable performance speed-ups. By embedding the recency-weighted average Reinforcement Learning method into a constraint propagation algorithm to dynamically learn the best operators, we show that it is possible to obtain robust algorithms with reliable performances on a range of sparse problems. Using a simple heuristic to compute initial weights, we also achieve significant performance speed-ups for dense problems.