求解拆卸线平衡问题的一种变邻域搜索算法

针对拆卸线平衡问题特点,建立以最小化工作站数、平滑指数、危害指数和需求指数为优化目标的多目标数学模型,进而提出一种改进的变邻域搜索算法求解该问题。所提算法采用了一种启发式方法产生初始解,并构造了3种邻域结构,扩大算法搜索范围。采用初始解集进行局部搜索,搜索过程使用一步改进策略,并引入禁忌表方法,提高算法全局寻优能力。通过对大量不同规模测试问题进行算法实验,并与现有方法对比,结果表明,所提变邻域搜索算法在求解效率与求解质量上具有优越性。     

一类解决Job Shop问题的禁忌搜索算法

针对Job shop问题，设计了一种改进的禁忌搜索算法（MTS算法）。MTS算法从多个初始解开始，将传统禁忌搜索算法由串行搜索结构变为并行搜索结构；采用互换和交叉两种邻域搜索函数，既有利于新邻域的探索又有利于交换信息；基于目标值的禁忌表保证了群体的多样性。实验表明，MTS算法克服了传统禁忌搜索算法的缺陷，具有较高的求解质量和鲁棒性。     

基于改进禁忌搜索算法的单机成组作业调度

提出了单机作业调度的改进禁忌搜索算法。以总流程时间最短为优化目标的具有作业分类与机器设置时间的单机成组作业高度问题是NP难问题,此问题的多项式求解方法不能保证求取最优解。基于改进禁忌搜索算法,建立了具有提交时间与分类设置时间的单机成组作业调度模型,可搜索到该问题的最优解。仿真实验表明,改进禁忌搜索算法能够搜索到最优解,其性能优于WSPT启发式算法,并且运行时间短,可应用于大规模调度问题。     

船舶布管系统的结构设计及自动化算法

为解决船舶管路自动布置问题,提出一种可扩展、模块化的软件结构,并给出软件系统的应用流程。通过对布置空间做预处理以减小问题规模,以Lee算法和禁忌搜索算法为例说明了生成管路和确定布管顺序的方法。对Lee算法做适应性修改,使其适合于三维管路生成且满足布置约束,结合船舶管路布置特点,为禁忌搜索算法设计解的通用形式、评价方法、禁忌策略和邻域结构,提高了算法的适应性、速度和收敛效果。通过实例验证了自动布置方法的有效性,并给出了对禁忌搜索算法的配置建议。     

求解置换调度问题的改进混合遗传算法

构造了求解极小化总完工时间的置换调度问题的改进混合遗传算法：先采用构造型启发式算法和随机方法共同产生初始种群，然后在选择、交叉和变异等遗传操作之前借助禁忌搜索算法寻找每个个体的局部最优解组成当前种群，再应用种群整体替换策略保存种群中的优秀个体构成新一代种群。改进混合遗传算法有机地结合了禁忌搜索算法的局部搜索性能和遗传算法的全局搜索性能。仿真实验表明，改进混合遗传算法具有比构造型启发式算法和禁忌搜索算法更好的鲁棒性和寻优性能。     

应用混合蚁群算法求解模糊作业车间调度问题

为解决蚁群算法求解时间过长和易陷入局部最优的问题,提出了一种求解模糊作业车间调度问题的混合算法,该算法将蚁群算法用于全局搜索.为了提高搜索效率,根据作业车间调度问题解的特征,提出一种基于关键工序的邻域搜索方法,并使用此邻域搜索方法的禁忌搜索算法嵌入蚁群算法.利用禁忌搜索算法较强的局部搜索能力,提高了蚁群算法的优化能力,改善了作业车间调度问题解的质量.实验结果验证了该混合搜索算法的有效性,其优化效果优于并行遗传算法和禁忌搜索算法.     

飞机复合材料部件装配调度算法研究

以优化飞机复合材料部件装配顺序、缩短部件装配时间为目标,构建了复合材料部件装配调度模型,提出了一种启发式算法与禁忌搜索算法相结合的调度算法（HTA-AJSP）,并对其进行了优化,通过采用变邻域搜索和禁忌规则,很好地避免了算法搜索陷入局部最优。通过分析与实例验证,并与启发式算法和局部搜索算法进行比较,所提出的算法在产品装配调度优化问题上取得了比较满意的效果,缩短了装配周期,提高了装配效率。     

飞机复合材料部件装配调度算法研究

以优化飞机复合材料部件装配顺序、缩短部件装配时间为目标,构建了复合材料部件装配调度模型,提出了一种启发式算法与禁忌搜索算法相结合的调度算法(HTA-AJSP),并对其进行了优化,通过采用变邻域搜索和禁忌规则,很好地避免了算法搜索陷入局部最优。通过分析与实例验证,并与启发式算法和局部搜索算法进行比较,所提出的算法在产品装配调度优化问题上取得了比较满意的效果,缩短了装配周期,提高了装配效率。     

基于混合禁忌搜索算法的供应链排序问题

分析非标准件加工企业供应链的特点,提出协同优化订单分配、生产调度和批量运输调度的多工厂多客户供应链排序问题。以工件的最长订货提前期与总成本加权之和最小化为目标,构建问题的数学模型。在分析解的最优性条件基础上,设计一种基于矢量组编码方法的混合禁忌搜索算法。算法对可行域进行分区,通过基于插入、交换两种邻域操作的禁忌搜索算法选择子区域,采用基于块结构邻域操作的禁忌搜索算法搜索子区域中的优良解。采用所提混合禁忌搜索算法对算例进行优化求解,并对采用不同编码方法、不同启发式算法的算例结果进行比较,结果表明所提出算法的有效性。     

带模糊预约时间的车辆路径问题的多目标禁忌搜索算法

为优化具有模糊预约时间的车辆路径问题,应用模糊事件给出了车队服务满意度的一个新的度量方法和求最大满意度的计算方法.建立了多目标数学规划模型,并提出多目标禁忌搜索算法求解Pareto最优解.采用随机车辆配载方法生成初始解放入候选解池中,提出插人可行邻域和2-Opt可行邻域进行邻域搜索.对池中的Pareto解进行并行的禁忌搜索得到局部Pareto解再注人池中,最后求得一组Pareto解.通过Solomon的benchmark算例,与非支配排序遗传算法Ⅱ进行对比实验,说明了所提算法的优越性.     

A Tabu Search Algorithm for Single Machine Scheduling with Release Times,Due Dates,and Sequence-Dependent Set-up Times

In this paper, we present a tabu search algorithm that schedules N jobs to a single machine in order to minimise the maximum lateness of the jobs. The release times, due dates, and sequence-dependent set-up times of the jobs are assumed to exist. We modified the original tabu search method to be suitable for the scheduling problem. The proposed tabu search algorithm is composed of two parts: a MATCS (modified apparent tardiness cost with set-ups) rule for finding an efficient initial solution, and the tabu search method to seek a near optimal solution from the initial solution. The experimental results show that the tabu search algorithm obtains much better solutions more quickly than the RHP (rolling horizon procedure) heuristic suggested by Ovacik and Uzsoy.     

Sequencing of an M machine flow shop with setup, processing and removal times separated

In this research, a flow shop scheduling problem in which setup, processing and removal times are separated, with the objective of minimizing makespan, is considered. A tabu search based heuristic is presented for solving the addressed problem. The proposed heuristic begins with the construction of artificial processing times for each operation; then a modified NEH algorithm is used to generate an initial solution, followed by a designed tabu search procedure applied for further improvement of the solution. The proposed heuristic, as well as the existing one, is evaluated in a large number of randomly generated problems. The results of the experimental investigations of the proposed heuristic algorithm and the existing heuristic in meeting the objective of makespan are also reported. It is found that the solution quality of the proposed tabu search heuristic is better than that of the existing heuristic, but for large size problems more computational efforts are needed; however, the CPU time is still acceptable.     

[供应链调度]分时电价和时变行程时间下的供应链调度

针对单机供应链调度问题,在生产计划和批次配送阶段分别考虑分时电价政策和时变行程时间。以总成本最小为目标建立混合整数规划模型,通过对模型的分析给出了最优解的性质,以此将模型分解为若干个批次的机器调度子问题。对于子问题优化,设计了子集划分启发式算法并证明了算法的最优性。对于主问题的优化,设计了自适应变邻域搜索算法。数值计算结果验证了模型和算法的有效性,证明了供应链集成调度能减少大量的能源消耗。     

A heuristic method for the combined location routing and inventory problem

The combined location routing and inventory problem (CLRIP) is used to allocate depots from several potential locations, to schedule vehicles’ routes to meet customers’ demands, and to determine the inventory policy based on the information of customers’ demands, in order to minimize the total system cost. Since finding the optimal solution(s) for this problem is a nonpolynomial (NP) problem, several heuristics for searching local optima have been proposed. However, the solutions for these heuristics are trapped in local optima. Global search heuristic methods, such as tabu search, simulated annealing method, etc., have been known for overcoming the combinatorial problems such as CLRIP, etc. In this paper, the CLRIP is decomposed into two subproblems: depot location-allocation problem, and routing and inventory problem. A heuristic method is proposed to find solutions for CLRIP. First of all, an initial solution for CLRIP is determined. Then a hybrid heuristic combining tabu search with simulated annealing sharing the same tabu list is used to improve the initial solution for each subproblem separately and alternatively. The proposed heuristic method is tested and evaluated via simulation. The results show the proposed heuristic method is better than the existing methods and global search heuristic methods in terms of average system cost.     

自动化制造单元最小完工时间调度问题的混合启发式算法

自动化制造单元最小完工时间调度问题属于NP-hard难题,目前尚缺乏有效的调度方法。为此,提出基于遗传和禁忌搜索的混合启发式算法,用以搜索一组最满意的机器人搬运作业排序。以遗传算法为基本结构,在初始种群产生和交叉、变异操作中引入禁忌搜索技术,以提高优化质量。基于搬运作业规则的初始种群构造算法和两阶段交叉、变异算子克服了传统算子对可行搬运作业排序的破坏,而邻域移动算子则保证了禁忌搜索的多样性和集中性。最后,随机实验结果验证了算法的有效性。     

Bi-objective heuristics for multi-item freights distribution planning problem in crossdocking networks

Crossdocking is a distribution strategy in which distribution centers operate as transfer points to harmonize the continuous physical flow through supply chain with the least storage. The coordination of this system in a network of crossdocks with multitype and time-restricted pickups and deliveries requires a complex planning. This paper addresses the distribution planning problem of crossdocking network, considering transshipment possibility among crossdocks and tardiness permission for some pickups. The problem is formulated as a bi-objective integer-programming model minimizing total transportation and holding costs and total tardiness. An innovative heuristic procedure to construct an effective initial solution and three improving algorithms based on appropriate meta-heuristics (variable neighborhood search, tabu search, and simulated annealing) is developed to solve the problem. Experimental results demonstrate the potency of proposed algorithms achieving remarkable results.     

A population-based hybrid ant system for quadratic assignment formulations in facility layout design

The facility layout design problem is an extensively studied research problem and belongs to nonpolynomial hard (NP-hard) combinatorial optimization problem. Quadratic assignment problem (QAP) is one of the formulations that is investigated for facility layout design because of its wide applicability. Ant colony optimization (ACO), a biologically inspired heuristic has centered on solving the QAP by achieving approximation as good as possible. This paper presents a population-based hybrid ant system (PHAS), which is an extension of the hybrid ant system (HAS) in which the size of the ant colony has been fixed. The performance of the proposed ant algorithm for QAP is compared with the existing metaheuristic implementations such as tabu search, reactive tabu search, simulated annealing, genetic hybrid method, HAS, and max–min ant system. The experimental results show that the proposed PHAS perform significantly better than the other existing algorithms of QAP.     

Scheduling jobs on parallel machines: a restricted tabu search approach

Many real scheduling problems are often much more complex than problems that are analytically tractable. The main difficulty in obtaining optimal job schedules arises from the existence of sequence dependent setup times among jobs and job release times. In this paper, we present a restricted tabu search algorithm that schedules jobs on parallel machines in order to minimise the maximum lateness of the jobs. The jobs have release times and due dates, and sequence-dependent setup times exist between the jobs. The parallel machines are either identical or non-identical in terms of the processing times of the jobs. The restricted tabu search algorithm employs a restricted search with the elimination of non-effective job moves, for finding the best neighbourhood schedule. The restricted search algorithm reduces search effort significantly while obtaining good quality final schedule. The experimental results show that the proposed algorithm obtains much better solutions more quickly than other heuristic algorithms such as the Rolling Horizon Procedure (RHP) heuristic, the basic tabu search and simulated annealing.     

The tricriteria flowshop scheduling problem

Multicriteria flowshop scheduling problems have been one of the most attractive subjects in recent years. In the multicriteria flowshop scheduling literature, a very limited number of studies have been performed on problems which include a tardiness criterion. In this paper a multicriteria (tricriteria) two-machine flowshop scheduling problem with a tardiness criterion is tackled. The objective is to minimise a weighted sum of total completion time, total tardiness and makespan. An integer programming model is proposed for the problem which belongs to NP-hard class. The modified NEH (Nawaz, Enscore and Ham) algorithm, a tabu search-based heuristic method, random search and the EDD rule (the earliest due date rule) are used to solve problems with up to 2,500 jobs. A computational analysis is conducted to evaluate the performance of the heuristics. The analysis shows that the heuristics are quite efficient, and the performance of the tabu search based heuristic is the best of all in terms of solution quality.