应用强化学习算法求解置换流水车间调度问题

面对日益增长的大规模调度问题,新型算法的开发越显重要.针对置换流水车间调度问题,提出了一种基于强化学习Q-Learning调度算法.通过引入状态变量和行为变量,将组合优化的排序问题转换成序贯决策问题,来解决置换流水车间调度问题.采用所提算法对OR-Library提供Flow-shop国际标准算例进行测试,并与已有的一些算法对比,结果表明算法的有效性.     

面向多目标流水车间调度的多种群多目标遗传算法

针对制造型企业普遍存在的流水车间调度问题,建立了以最小化最迟完成时间和总延迟时间为目标的多目标调度模型,并提出一种基于分解方法的多种群多目标遗传算法进行求解.该算法将多目标流水车间调度问题分解为多个单目标子问题,并分阶段地将这些子问题引入到算法迭代过程进行求解.算法在每次迭代时,依据种群的分布情况选择各子问题的最好解及与其相似的个体分别为当前求解的子问题构造子种群,通过多种群的进化完成对多个子问题最优解的并行搜索.通过对标准测试算例进行仿真实验,结果表明所提出的算法在求解该问题上能够获得较好的非支配解集.     

直觉模糊集相似度遗传算法求解多目标车间调度问题

为了提高高维多目标置换流水车间调度问题的求解质量,提出基于直觉模糊集相似度的遗传算法(similarity of intuitionistic fuzzy sets GA,SIFS_GA).算法中分别将参考解和Pareto解映射为参考解直觉模糊集和Pareto解直觉模糊集.计算两个集合之间的直觉模糊相似度,用以判断Pareto解的优劣.以直觉模糊集相似度值引导多目标遗传算法进化.对6个CEC标准测试集与10个流水车间调度测试实例进行仿真实验,结果表明SIFS_GA算法性能优于常用的多目标优化算法,且可以有效解决多目标置换流水车间调度问题,尤其在解决规模较大的问题上是一种有效方法.     

Solving a bi-criteria permutation flow-shop problem using shuffled frog-leaping algorithm

Flow shop problems as a typical manufacturing challenge have gained wide attention in academic fields. In this paper, we consider a bi-criteria permutation flow shop scheduling problem, where the weighted mean completion time and the weighted mean tardiness are to be minimized simultaneously. Due to the complexity of the problem, it is very difficult to obtain optimum solution for this kind of problems by means of traditional approaches. Therefore, a new multi-objective shuffled frog-leaping algorithm (MOSFLA) is introduced for the first time to search locally Pareto-optimal frontier for the given problem. To prove the efficiency of the proposed algorithm, various test problems are solved and the reliability of the proposed algorithm, based on some comparison metrics, is compared with three distinguished multi-objective genetic algorithms, i.e. PS-NC GA, NSGA-II, and SPEA-II. The computational results show that the proposed MOSFLA performs better than the above genetic algorithms, especially for the large-sized problems.     

A biased‐randomized iterated local search for the distributed assembly permutation flow‐shop problem

Modern production systems require multiple manufacturing centers—usually distributed among different locations—where the outcomes of each center need to be assembled to generate the final product. This paper discusses the distributed assembly permutation flow‐shop scheduling problem, which consists of two stages: the first stage is composed of several production factories, each of them with a flow‐shop configuration; in the second stage, the outcomes of each flow‐shop are assembled into a final product. The goal here is to minimize the makespan of the entire manufacturing process. With this objective in mind, we present an efficient and parameter‐less algorithm that makes use of a biased‐randomized iterated local search metaheuristic. The efficiency of the proposed method is evaluated through the analysis of an extensive set of computational experiments. The results show that our algorithm offers excellent performance when compared with other state‐of‐the‐art approaches, obtaining several new best solutions.     

A water-flow algorithm for flexible flow shop scheduling with intermediate buffers

We investigate the flexible flow shop scheduling problem with limited or unlimited intermediate buffers. A common objective of the problem is to find a production schedule that minimizes the completion time of jobs. Other objectives that we also consider are minimizing the total weighted flow time of jobs and minimizing the total weighted tardiness time of jobs. We propose a water-flow algorithm to solve this scheduling problem. The algorithm is inspired by the hydrological cycle in meteorology and the erosion phenomenon in nature. In the algorithm, we combine the amount of precipitation and its falling force to form a flexible erosion capability. This helps the erosion process of the algorithm to focus on exploiting promising regions strongly. To initiate the algorithm, we use a constructive procedure to obtain a seed permutation. We also use an improvement procedure for constructing a complete schedule from a permutation that represents the sequence of jobs in the first stage of the scheduling problem. To evaluate the proposed algorithm, we use benchmark instances taken from the literature and randomly generated instances of the scheduling problem. The computational results demonstrate the efficacy of the algorithm. We have also obtained several improved solutions for the benchmark instances using the proposed algorithm. We further illustrate the algorithm’s capability for solving problems in practical applications by applying it to a maltose syrup production problem.     

置换流水车间调度的蚁群优化算法

提出了一种求解置换流水车间调度的蚁群优化算法。该算法的要点是结合了NEH启发式算法和蚁群优化方法。理论论证和对置换流水车间调度问题的基准测试表明了该算法的有效性。     

求解流水线调度问题的万有引力搜索算法

研究了以最大完工时间为目标的流水线调度问题,使用万有引力算法求解调度问题,提出了一种最大排序规则,利用物体间各个位置分量值存在的大小次序关系,并结合随机键编码的方法产生,将物体的连续位置转变成了一个可行的调度方案;提出了一种边界变异的策略使得越界的物体不再聚集在边界上,而是分布在边界附近的可行空间内,从而增加种群的多样性;结合交换算子和插入算子提出了一种新的局部搜索算法,有效地避免了算法陷入局部最优值,进一步提高了解的质量.最后证明了算法的收敛性,并且计算了算法的时间复杂度和空间复杂度,仿真实验说明了所得算法的有效性.     

Multi-objective adaptive large neighborhood search for distributed reentrant permutation flow shop scheduling

Factory management plays an important role in improving the productivity and quality of service in the production process. In particular, the distributed permutation flow shop scheduling problem with multiple factories is considered a priority factor in the factory automation. This study proposes a novel model of the developed distributed scheduling by supplementing the reentrant characteristic into the model of distributed reentrant permutation flow shop (DRPFS) scheduling. This problem is described as a given set of jobs with a number of reentrant layers is processed in the factories, which compromises a set of machines, with the same properties. The aim of the study is to determine the number of factory needs to be used, jobs assignment to certain factory and sequence of job assigned to the factory in order to simultaneously satisfy three objectives of minimizing makespan, total cost and average tardiness. To do this, a novel multi-objective adaptive large neighborhood search (MOALNS) algorithm is developed for finding the near optimal solutions based on the Pareto front. Various destroy and repair operators are presented to balance between intensification and diversification of searching process. The numerical examples of computational experiments are carried out to validate the proposed model. The analytical results on the performance of proposed algorithm are checked and compared with the existing methods to validate the effectiveness and robustness of the proposed potential algorithm in handling the DRPFS problem.     

基于蝙蝠算法的PFSP调度干扰管理研究

对置换流水车间调度问题分别加入单工件、双工件、四工件干扰因素,采用较新颖的蝙蝠算法对其进行优化,并同较成熟的粒子群算法进行比较,结果表明新型启发式智能蝙蝠智能算法性能不仅仅适用于生产调度干扰管理问题的求解且优于基本粒子群算法,同时对其加权系数进行测试分析,结果表明了加权系数对生产调度干扰管理影响的权重。     

基于Hopfield神经网络的作业车间生产调度方法

该文提出了基于Hopfield神经网络的作业车间生产调度的新方法.文中给出了作业车 间生产调度问题(JSP)的约束条件及其换位矩阵表示,提出了新的包括所有约束条件的计算能 量函数表达式,得到相应的作业车间调度问题的Hopfield神经网络结构与权值解析表达式,并 提出相应的Hopfield神经网络作业车间调度方法.为了避免Hopfield神经网络容易收敛到局部 极小,从而产生非法调度解的缺点,将模拟退火算法应用于Hopfield神经网络求解,使Hopfield 神经网络收敛到计算能量函数的最小值0,从而保证神经网络输出是一个可行调度方案.该文 改进了已有文献中提出的作业调度问题的Hopfield神经网络方法,与已有算法相比,能够保证 神经网络稳态输出为可行的作业车间调度方案.     

Minimizing makespan for flow shop scheduling problem with intermediate buffers by using hybrid approach of artificial immune system

In this study, three new meta-heuristic algorithms artificial immune system (AIS), iterated greedy algorithm (IG) and a hybrid approach of artificial immune system (AIS-IG) are proposed to minimize maximum completion time (makespan) for the permutation flow shop scheduling problem with the limited buffers between consecutive machines. As known, this category of scheduling problem has wide application in the manufacturing and has attracted much attention in academic fields. Different from basic artificial immune systems, the proposed AIS-IG algorithm is combined with destruction and construction phases of iterated greedy algorithm to improve the local search ability. The performances of these three approaches were evaluated over Taillard, Carlier and Reeves benchmark problems. It is shown that the AIS-IG and AIS algorithms not only generate better solutions than all of the well-known meta heuristic approaches but also can maintain their quality for large scale problems.     

基于约束理论的Flow-shop分解协调算法

There are many flow shop problems of throughput (denoted by FSPT) with constraints of due date in real production planning and scheduling. In this paper, a decomposition and coordination algorithm is proposed based on the analysis of FSPT and under the support of TOC (theory of constraint). A flow shop is at first decomposed into two subsystems named PULL and PUSH by means of bottleneck. Then the subsystem is decomposed into single machine scheduling problems, so the original NP-HARD problem can be transferred into a serial of single machine optimization problems finally. This method reduces the computational complexity, and has been used in a real project successfully.     

Elucidating multiprocessors flow shop scheduling with dependent setup times using a twin particle swarm optimization

Particle swarm optimization (PSO) is a novel metaheuristic, which has been applied in a wide variety of production scheduling problems. Two basic characteristics of this algorithm are its efficiency and effectiveness in providing high-quality solutions. In order to improve the traditional PSO, this study proposes the incorporation of a local search heuristic into the basic PSO algorithm. The new, hybrid, metaheuristic is called “twin particle swarm optimization (TPSO)”. The proposed metaheuristic scheme is applied to a flow shop with multiprocessors scheduling problem, which can be considered a real world case regarding the production line. This study, as far as the multiprocessors flow shop production system is concerned, utilizes sequence dependent setup times as constraints. Finally, simulated data confirm the effectiveness and robustness of the proposed algorithm. The data test results indicate that TPSO has potential to replace PSO and become a significant heuristic algorithm for similar problems.     

An effective neighborhood search algorithm for scheduling a flow shop of batch processing machines

This paper considers scheduling problem of flow shop with many batch processing machines and objective of maximum lateness. An effective neighborhood search algorithm (NSA) is proposed for the problem, in which a job permutation and a batch permutation are used to indicate the solution of two sub-problems, respectively. Each job permutation consists of several family-permutations for the representation of jobs from the same family. Two swaps are applied to two permutations to produce new solutions. NSA is applied to a number of instances and compared with some methods, and computational results validate the good performance of NSA.     

A Decomposition and Coordination Scheduling Method for Flow-shop Problem Based on TOC

There are many flow shop problems of throughput (denoted by FSPT) with constraints of due date in real production planning and scheduling. In this paper, a decomposition and coordination algorithm is proposed based on the analysis of FSPT and under the support of TOC (theory of constraint). A flow shop is at first decomposed into two subsystems named PULL and PUSH by means of bottleneck. Then the subsystem is decomposed into single machine scheduling problems, so the original NP-HARD problem can be transferred into a serial of single machine optimization problems finally. This method reduces the computational complexity, and has been used in a real project successfully.     

An efficient local search for minimizing completion time variance in permutation flow shops

This paper illustrates that by exploiting the structure of hard combinatorial optimization problems, efficient local search schemes can be designed that guarantee performance in solution quality and computational time. A two-phase local search algorithm is developed and applied to the permutation flow shop scheduling problem, with the objective of minimizing the completion time variance. New and significant analytical insights necessary for effectively solving the permutation flow shop problem are also presented and used in this research. Computational results indicate that for test problems, the local search obtained optimal solutions for many instances, within few seconds of CPU time. For other benchmark problems with jobs between 50 and 100, the proposed algorithm, ADJ-Reduced improved the existing best known values within a practical time frame.     

求解置换流水线调度问题的改进萤火虫优化算法

针对最小化最大完成时间的置换流水线调度问题,提出了一种改进的离散萤火虫优化算法。在传统萤火虫优化算法的基础上,采用基于升序排序的随机键编码方式对萤火虫种群进行离散化处理,使用NEH算法对萤火虫种群进行初始化处理,结合遗传算法的交叉变异思想改进位置更新策略,采用个体变异方式解决孤立个体问题,提高算法的寻优能力。最后通过典型算例对改进算法进行仿真测试,实验结果表明该算法求解置换流水线调度问题时具备很强的寻优能力和鲁棒性,明显优于传统萤火虫优化算法和遗传算法,是解决置换流水线调度问题的一种有效算法。     

解决具有混合约束柔性流水车间调度问题的粒子群优化算法

针对既存在阻塞限制工件又存在无等待约束工件的柔性流水车间调度问题, 提出了一种离散粒子群优化的求解方法。该方法采用基于排列的编码形式, 设计了推进—迭代算法进行解码并计算问题目标值, 利用离散粒子群优化算法进行全局优化, 利用迭代贪婪(iterated greedy, IG)算法提高种群个体的局部搜索能力。此外, 根据问题特点, 提出最早释放优先(first release first, FRF)和最早完工优先(first complete first, FCF)两种机器分配策略。仿真结果表明, 所提出的方法求解混合约束下柔性流水车间调度问题是可行的、有效的。     

A dual encoding-based meta-heuristic algorithm for solving a constrained hybrid flow shop scheduling problem

Though scheduling problems have been largely investigated by literature over the last 50 years, this topic still influences the research activity of many experts and practitioners, especially due to a series of studies which recently emphasized the closeness between theory and industrial practice. In this paper the scheduling problem of a hybrid flow shop with m stages, inspired to a truly observed micro-electronics manufacturing environment, has been investigated. Overlap between jobs of the same type, waiting time limit of jobs within inter-stage buffers as well as machine unavailability time intervals represent just a part of the constraints which characterize the problem here investigated. A mixed integer linear programming model of the problem in hand has been developed with the aim to validate the performance concerning the proposed optimization technique, based on a two-phase metaheuristics (MEs). In the first phase the proposed ME algorithm evolves similarly to a genetic algorithm equipped with a regular permutation encoding. Subsequently, since the permutation encoding is not able to investigate the overall space of solutions, a random search algorithm equipped with an m-stage permutation encoding is launched for improving the algorithm strength in terms of both exploration and exploitation. Extensive numerical studies on a benchmark of problems, along with a properly arranged ANOVA analysis, demonstrate the statistical outperformance of the proposed approach with respect to the traditional optimization approach based on a single encoding. Finally, a comprehensive comparative analysis involving the proposed algorithm and several metaheuristics developed by literature demonstrated the effectiveness of the dual encoding based approach for solving HFS scheduling problems.