考虑机器检修的热轧钢管批量计划方法

本文从无缝钢管生产实际中提取并定义了周期性机器检修环境下的钢管热轧批量计划问题,基于无缝钢管生产的特殊性,将该问题抽象为一类考虑机器检修和机器调整时间的单机调度问题,并建立了以最小化机器闲置和机器调整时间为目标的数学模型.针对批量间的机器调整时间取决于钢管规格的变化这一特性,提出了最小调整时间排序规则,证明了该规则在不考虑检修计划时具有最优性.进而,以此为基础建立了循环求解框架,并设计了两阶段启发式算法.基于实际生产数据设计了多种问题规模的实验,验证了算法的有效性,并从实际应用角度对结果进行了分析.     

顺序依赖的调整时间和拖期的无缝钢管热轧批量调度算法

无缝钢管热轧生产存在一类特殊的顺序依赖机器调整时间,调整时间依赖于相邻轧制批量间的规格切换,与批量间规格呈线性函数关系.针对具有此类调整时间的热轧批量调度问题,进一步考虑交货期要求,探讨了调整时间与交货期之间的性质特征,并以最小化总机器调整时间和最小化总拖期为目标,基于进化算法框架设计了快速重排序邻域搜索多目标算法(fast reordering neighborhood search multi-objective algorithm,FRNSM).在算法初始阶段,提出规格优化规则和交货时间优化规则,对初始解进行局部优化;邻域搜索阶段设计了基于同交货时间规格优化的快速重排序邻域搜索策略,以优化机器调整时间和拖期;针对问题多目标特征,基于Pareto思想设计解的接受准则.基于实际生产数据的实验结果表明,FRNSM算法对此类问题具有很好的求解效果.     

基于最小机器数的柔性作业车间调度研究

针对期望以最小机器数完成生产的柔性作业车间调度问题,建立了最小化最大完工时间为内层目标,最小机器数为外层目标的双层优化模型,即在满足交货期、最小化最大完工时间的条件下,尝试减少机器数量,以寻求车间调度的最少机器数.依据模型、算法特点,设计了一种基于大变异策略的遗传算法,该算法采用二维染色体编码、顺序选择策略,同时运用优...     

改进离散粒子群算法求解柔性流水车间调度问题

针对以最小化完工时间为目标的柔性流水车间调度问题(FFSP),提出了一种改进离散粒子群(DPSO)算法.所提算法重新定义粒子速度和位置的相关算子,并引入编码矩阵和解码矩阵来表示工件、机器以及调度之间的关系.为了提高柔性流水车间调度问题求解的改进离散粒子群算法的初始群体质量,通过分析初始机器选择与调度总完工时间的关系,首次提出一种基于NEH算法的最短用时分解策略算法.仿真实验结果表明,该算法在求解柔性流水车间调度问题上有很好的性能,是一种有效的调度算法.     

基于PCTSP的热轧单元计划模型与算法

根据钢铁企业热轧产品生产工艺约束条件,将热轧生产轧制单元计划模型归结为奖金收集旅行商问题,设计了蚁群最优化算法对模型进行求解.引用某钢铁企业热轧生产轧制单元计划编制的实际问题对模型和算法进行了验证,并与遗传算法的求解结果进行了对比.实验结果表明模型和算法的优化效果和时间效率都是令人满意的.该模型和算法经过改进后可应用到包含多个轧制单元计划的轧制批量计划优化问题中.     

基于混合遗传算法的热轧生产调度优化方法

热轧生产调度是一个复杂的约束组合优化问题,其生产约束包括连续轧制板坯的宽度、厚度和硬度跳变要求,轧制单元的最大长度,产品库存及交货期等.基于多旅行商模型,建立了热轧生产批量调度问题的优化模型,并提出一种混合遗传算法(遗传算法、局部搜索)求解该问题.通过应用串行边重组和并行边重组的遗传交叉算子,算法在优化过程中可以很好地处理调度约束.针对工业数据的仿真结果证明该调度模型和混合遗传算法的并行求解策略可以有效地解决热轧生产批量调度问题.     

基于混合遗传算法的热轧生产调度优化方法

热轧生产调度是一个复杂的约束组合优化问题，其生产约束包括连续轧制板坯的宽度、厚度和硬度跳变要求，轧制单元的最大长度，产品库存及交货期等。基于多旅行商模型，建立了热轧生产批量调度问题的优化模型，并提出一种混合遗传算法（遗传算法、局部搜索）求解该问题。通过应用串行边重组和并行边重组的遗传交叉算子，算法在优化过程中可以很好地处理调度约束。针对工业数据的仿真结果证明该调度模型和混合遗传算法的并行求解策略可以有效地解决热轧生产批量调度问题。     

基于改进双档案多目标进化算法的柔性作业车间批量流混排调度

针对柔性作业车间批量流调度问题,基于快速换模技术,考虑可变子批划分柔性、子批混排加工柔性、自动换模柔性和物料运输柔性,建立以最小化完工时间和加工子批总数为目标的混排调度优化模型,并提出一种改进双档案多目标进化算法以优化目标函数。基于进化算法框架,设计了基于超体积指标和基于改进帕累托支配的双档案筛选机制,以平衡种群的收敛性和多样性;针对批量流混排调度问题特征,在解码阶段提出正/逆解码和子批拆分左移策略,在邻域探索和全局搜索阶段分别设计子批划分和混排调度的自适应进化算子,以提高算法的全局搜索与局部搜索能力。基于不同规模算例,测试了提出算法与经典多目标算法的性能。实验结果表明,该算法在收敛性与多样性上具有明显优势。     

混合离散人工蜂群算法求解含不相关并行机的分布式柔性流水线调度

研究每阶段含不相关并行机的分布式柔性流水线调度问题.考虑顺序相关准备时间和工件动态到达时间,以最小化总加权提前/拖期惩罚为目标建立整数规划模型,提出一种融合离散差分进化算法、变邻域下降算法和局域搜索的混合离散人工蜂群算法以获取近优解.该算法采用基于工厂-工件号的编码以及基于机器最早空闲时间的动态解码机制,通过随机规则和均衡分派策略生成初始工厂-工件序列群,在引领蜂阶段引入离散差分进化算法产生优质工厂-工件序列,在跟随蜂阶段利用变邻域下降算法在被选择序列附近继续搜索以得到邻域序列,在侦察蜂阶段设计基于关键/非关键工厂间插入的局域搜索提高算法搜索能力.通过仿真实验测试不同规模的算例,实验结果表明,所提出的混合离散人工蜂群算法表现出较好的求解性能.     

改进的粒子群算法在多目标车间调度的应用

利用粒子群算法解决车间调度问题,是一种有效的策略。对粒子群算法进行分析,针对多目标的柔性车间调度问题,构建了以加工时间最小化、加工成本最小化和单机器最大负荷最小化的多目标柔性车间调度模型。提出基于交叉变异的变参粒子群算法,以提高其跳出局部最优快速达到全局最优的能力。同时,引入智能小车概念,将运输时间考虑到此调度中。并将该方法用于某离散制造业的柔性车间作业调度中,最后验证了该算法的实用性及高效性。     

Algorithms for job scheduling problems with distinct time windows and general earliness/tardiness penalties

This paper addresses the single machine scheduling problem with distinct time windows and sequence-dependent setup times. The objective is to minimize the total weighted earliness and tardiness. The problem involves determining the job execution sequence and the starting time for each job in the sequence. An implicit enumeration algorithm denoted IE and a general variable neighborhood search algorithm denoted GVNS are proposed to determine the job scheduling. IE is an exact algorithm, whereas GVNS is a heuristic algorithm. In order to define the starting times, an O(n²) idle time insertion algorithm (ITIA) is proposed. IE and GVNS use the ITIA algorithm to determine the starting time for each job. However, the IE algorithm is only valid for instances with sequence-independent setup times, and takes advantage of theoretical results generated for this problem. Computational experiments show that the ITIA algorithm is more efficient than the only other equivalent algorithm found in the literature. The IE algorithm allows the optimal solutions of all instances with up to 15 jobs to be determined within a feasible computational time. For larger instances, GVNS produces better-quality solutions requiring less computational time compared with the other algorithm from the literature.     

Scheduling in a manufacturing shop with sequence-dependent setups

In this paper, the problem of scheduling multiple jobs in a flexible manufacturing cell with multiple machine stations is addressed. Due to the large capital investments that usually characterize flexible manufacturing systems (FMS), an area of control of great interest to system users is that of maximizing the system performance through the minimization of machine idle and setup times. The magnitude of total time spent on machine setups and idle times is influenced by the availability of jobs, job mix, similarities of jobs and job scheduling procedure used. Similar jobs on the same machine require less setup times. Similarly, the use of an adequate scheduling method also reduces total idle and setup times. Such reduction improves the flow times of jobs. In this paper, a heuristic algoritm for scheduling jobs with sequence dependent setup times in a FMS is presented. The measure of performance for evaluating schedule adequacy is the production makespan.     

Scheduling with dependent setups and maintenance in a textile company

This paper deals with a scheduling problem with machine maintenance in a textile company. In the production system, the sequence-dependent setup time of a job cannot be ignored when a switch between two different jobs occurs. This study presents a heuristic to minimize the completion time, or equivalently the total setup time subject to periodic maintenance and due dates. The performance of the heuristic is evaluated by comparing its solution with the solution of the branch-and-bound algorithm. The real data are used to demonstrate the effectiveness of the heuristic.     

并行机成组调度问题的启发式算法

研究了优化目标为总拖后/提前时间最小化的并行机成组调度问题,提出了一种三阶段启发式近似求解算法。首先把并行机问题看成单机问题,以最小化总拖后时间为优化目标排列工件的加工次序;然后将工件按第一阶段所求得的次序指派到最先空闲的并行的机器上;最后采用改进的GTW算法对各机器上的工件调度插入适当的空闲时间。计算表明该算法能够在很短的时间内给出大规模调度问题的近似最优解。     

A variable neighborhood search for minimizing total weighted tardiness with sequence dependent setup times on a single machine

This paper deals with the single machine scheduling problem to minimize the total weighted tardiness in the presence of sequence dependent setup. Firstly, a mathematical model is given to describe the problem formally. Since the problem is NP-hard, a general variable neighborhood search (GVNS) heuristic is proposed to solve it. Initial solution for the GVNS algorithm is obtained by using a constructive heuristic that is widely used in the literature for the problem. The proposed algorithm is tested on 120 benchmark instances. The results show that 37 out of 120 best known solutions in the literature are improved while 64 instances are solved equally. Next, the GVNS algorithm is applied to single machine scheduling problem with sequence dependent setup times to minimize the total tardiness problem without changing any implementation issues and the parameters of the GVNS algorithm. For this problem, 64 test instances are solved varying from small to large sizes. Among these 64 instances, 35 instances are solved to the optimality, 16 instances' best-known results are improved, and 6 instances are solved equally compared to the best-known results. Hence, it can be concluded that the GVNS algorithm is an effective, efficient and a robust algorithm for minimizing tardiness on a single machine in the presence of setup times.     

面向人机物三元数据的热轧调度问题研究

随着钢铁行业的数字化发展,其订单逐渐趋于多样化和随机化,这对热轧调度模型的适应性和灵活性等提出了新的要求.针对热轧调度问题,当前的主流方法是启发式算法,但其存在两个问题:一是没有考虑数据的组织表示;二是此类算法具有很强的针对性,当问题发生很小的改变就需要进行复杂的参数调整.相比之下,机器学习具有更好的适应性和灵活性,对此,采用本体进行人机物三元数据的组织表示,提出一种指针网络$+$强化学习的热轧调度求解方法.采用指针网络来学习序列到序列的映射,同时为解决指针网络训练困难和性能不高等问题,通过actor-critic网络进行训练,提高模型的准确性和收敛速度.最后,通过设计相应的实验对算法的性能进行仿真并与LK-H的局部搜索算法进行对比,进一步验证了所提出方法的有效性.     

Solving a stochastic single machine problem with initial idle time and quadratic objective

We study a static single machine scheduling problem in which processing times are stochastic, due-dates and penalties for not completing jobs on time are deterministic, and an initial fixed idle time is allowed to be inserted before the processing of the first job begins on the machine. The objective is to determine the optimal sequence and the optimal initial idle time that jointly minimize the expected value of the sum of a quadratic cost function of idle time and the weighted sum of a quadratic function of job lateness. The problem is NP-hard to solve; however, we develop an exact algorithm based on a precedence relation structure among adjacent jobs. Our extensive computational results show that the algorithm can solve large problem instances quickly. We also demonstrate that the proposed problem is general in the sense that its special cases reduce to new stochastic models while its limiting cases simplify to some deterministic models.     

A numerical method in optimal production and setup scheduling ofstochastic manufacturing systems

In this paper, we consider optimal production and setup scheduling in a failure-prone manufacturing system consisting of a single machine. The system can produce several types-of products, but at any given time it can only produce one type of product. A setup is required if production is to be switched from one type of product to another. The decision variables are a sequence of setups and a production plan. The objective of the problem is to minimize the cost of setup, production, and surplus. An approximate optimality condition is given together with a computational algorithm for solving the optimal control problem