基于JIT的一类零排序问题的通用模型及解法

提出了在ＪＩＴ意义下由ｍ台加工设备组成的柔性制造系统一类零件排序问题的通用模型，证明了该问题是一个ＮＰ完全问题，构造出求解核模型的一个启发式算法。数值例子表明，该算法构思新颖，计算复杂性低，有较为理想的优化效果，是一个具有工程应用价值实用算法。     

一个基于JIT的FMS零件排序问题的模型及解法

本文提出了一个在ＪＩＴ意义下，由ｍ台加工设备组成的加工系统中零件排序的多目标优化模型，然后根据本文给出的单机问题新算法，构造出一个求解该模型的启发式算法，该算法的构思新颖，计算复杂性低，优化效果好，是一个具有工程实用价值的有效算法。最后给出了一个数值例子。     

基于全局图遍历的加权频繁模式挖掘算法

为解决加权遍历模式挖掘问题,提出了一种从边加权有向图到顶点加权有向图的变换模型.基于该模型,提出了基于全局图遍历加权频繁模式挖掘算法.在该算法中,利用图全局拓扑信息评估遍历模式的加权支持度,将剪枝问题转化为模式可扩展性问题,进而利用可扩展模式的向下闭合特性产生候选模式集.实验结果表明,该算法是一个高效的基于图遍历的加权频繁模式挖掘算法.     

具有同时集送货需求的车辆路径问题的自适应混合遗传算法

为满足配送货物和收集货物的需求,给出了具有同时集送货需求的车辆路径问题的混合整数规划模型,设计了求解该模型的自适应混合遗传算法。该算法以最优划分方法计算适应值,邻域搜索法作为变异算子,设计了新颖的交叉算子和群体更新策略,定义了群体多样性结构和变异概率的变化规律。通过仿真实验,并与已有优化算法比较,表明该算法是求解具有同时集送货需求的车辆路径问题和一般车辆路径问题的一个有效方法。     

论文简介

论文简介柔性制造零件排序的一个工业实用算法零件排序问题是柔性制造系统生产调度管理中静态作业计划的基本问题之一。本文通过对系统加工设备台数ｍ＝２的约翰逊算法的分析和研究，利用“插值”的思想和局部优化原理，对于ｍ≥３的零件排序问题，构造出一个新的启发式算...     

敏捷制造下合作伙伴选择的有效算法

为实现虚拟企业实践中有效的合作伙伴选择，根据制造成本和产品上市时间这两个重要因素建立了问题优化模型。基于此模型，给出了一个两阶段求解算法。该算法首先通过一个多项式的精确算法求得最早完成时间的解，保证了解的可行性。然后以最早完成时间作为可行初始解，利用一个启发式算法对解进行改进。该算法最终使得在完成时间的约束下制造成本最小。最后给出了一个应用案例说明了该算法的有效性。     

三维装载约束下带时间窗的车辆路径问题

为提高配送车辆的效率,集成研究了三维装载约束下带时间窗的车辆路径问题。提出了该问题的描述性模型,设计了一个混合禁忌搜索算法。该算法以空间装载算法、基础启发式算法和禁忌搜索算法为基础。针对测试数据集的计算结果表明,该算法有效地解决了三维装载约束下带时间窗的车辆路径问题。     

基于选择性集成算法的浸出率混合预测模型

一个精确的模型对浸出过程中浸出率预测是十分重要的。针对湿法冶金浸出生产过程中浸出率在线检测的难点,提出一种有效的混合建模方法,建立浸出率的预测模型。在对浸出过程进行分析后,建立了一个浸出过程机理模型。由于机理模型与实际之间存在着较大的误差,因此建立了混合模型来减少误差。随后,针对小样本建模问题,提出了基于二进制PSO算法的选择性bagging集成算法,并将该算法应用于混合模型的误差补偿中去。实验结果表明该混合模型的预测精度比其他模型的预测精度高。     

求解作业车间调度问题的快速启发式算法

首先将作业车间调度问题转换为一个搭积木模型,受这个直观模型的启发,提出了一个启发式的搭积木规则,该规则综合考虑了已经搭好的积木的顶高和将要搭积木的剩余高度。基于这个规则,提出了一个求解作业车间调度问题的快速启发式算法,对国际上通用的benchmark例的模拟实验结果表明,提出的算法优于经典的优先分配启发式算法。     

面向大规模定制生产的e-制造单元目标层解分析优化规划模型

e-制造单元规划是一个较为复杂的大型设计优化问题,全局优化性和算法复杂性的协调处理始终是解决该问题的重点和难点.通过深入分析e-制造单元的结构和特点,提出一种e-制造逻辑单元的双级式层次结构模型,并采用目标层解分析(ATC)方法建立一种e-制造逻辑单元协同规划模型以及相应的算法模型和解决方案.该模型针对大规模定制产品的制造问题,综合考虑各零件族在制造过程中的耦合关系,在零件聚类优化的基础上,实现e-制造单元的优化规划.通过深入细致的算法复杂度和收敛性分析,并结合具体的实例验证,论证该模型的合理性和有效性.     

A heuristic algorithm for solving flexible job shop scheduling problem

This paper deals with the flexible job shop scheduling problem with the objective of minimizing the makespan. An efficient heuristic based on a constructive procedure is developed to obtain high-quality schedules very quickly. The algorithm is tested on benchmark instances from the literature in order to evaluate its performance. Computational results show that, despite its simplicity, the proposed heuristic can obtain effective solutions in very short and nearly zero time and is comparable with even metaheuristic algorithms and promising for practical problems.     

A genetic-algorithm-based optimization model for scheduling flexible assembly lines

In this paper, a scheduling problem in the flexible assembly line (FAL) is investigated. The mathematical model for this problem is presented with the objectives of minimizing the weighted sum of tardiness and earliness penalties and balancing the production flow of the FAL, which considers flexible operation assignments. A bi-level genetic algorithm is developed to solve the scheduling problem. In this algorithm, a new chromosome representation is presented to tackle the operation assignment by assigning one operation to multiple machines as well as assigning multiple operations to one machine. Furthermore, a heuristic initialization process and modified genetic operators are proposed. The proposed optimization algorithm is validated using two sets of real production data. Experimental results demonstrate that the proposed optimization model can solve the scheduling problem effectively.     

A constructive heuristic for the integrated scheduling of machines and multiple-load material handling equipment in job shops

This paper extends the traditional job shop scheduling problem (JSP) by incorporating the routing and scheduling decisions of the material handling equipment. It provides a generic definition and a mixed integer linear programming model for the problem considering the case of heterogeneous multiple-load material handling equipment. A constructive heuristic is developed for solving the problem. This heuristic is based on the well-known Giffler and Thompson’s algorithm for the JSP with modifications that account for the routing decisions of the material handling equipment and their effect on the start times of the manufacturing operations. Different dispatching rules are integrated into the heuristic, and experiments are conducted to study their effect on the makespan along with the determination of the computational time requirements of the developed heuristic.     

[车间调度]并行多机开放车间调度问题的模型与算法

发光二极管制造过程中,晶粒分类拣选工序的调度问题是典型的并行多机开放车间调度问题,属于NP-hard问题。研究了该调度问题以最小化总加权完工时间为目标的求解模型与算法。根据问题特性构建了可获得最优解的混合整数规划模型,并设计了同时考虑质量与求解效率的启发式算法和改进粒子群优化算法。仿真结果显示,启发式算法和改进粒子群优化算法都能在合理的时间内迅速有效地获得较佳的调度解。     

基于遗传优化算法求解作业车间调度问题

将一类具有路径柔性的作业车间调度问题，分解为任务优化分配和任务优化调度两个子问题，结合遗传算法和启发式算法，提出了基于遗传算法的优化调度算法，并有实例证实了该算法的有效性。     

具有工件约束的模具制造优化调度算法研究

为解决具有工件约束的模具制造优化调度问题,提出了一种利用蚁群算法和优先分配启发式调度算法相结合的调度算法。该算法能够方便地描述问题的约束条件的特点。首先,由蚁群算法确定模具零件各工序所用的加工机床,用节点模式下的有向图描述问题的解空间,用蚂蚁种子信息素踪迹更新策略对信息素进行更新,以获得问题的解;然后,利用优先分配启发式调度算法确定在同一台机床上加工的各零件的先后顺序。实验结果验证了算法的有效性。     

A simulated annealing approach to makespan minimization on identical parallel machines

This paper addresses a makespan minimization scheduling problem on identical parallel machines. Several heuristic algorithms have been proposed to tackle the problem. In this paper, a very effective simulated annealing method is proposed to generate the near-optimal solution. Computational results demonstrate that the proposed heuristic is very accurate and that it outperforms the existing methods.     

A scheduling model for serial jobs on parallel machines with different preventive maintenance (PM)

In factories during production, preventive maintenance (PM) scheduling is an important problem in preventing and predicting the failure of machines, and most other critical tasks. In this paper, we present a new method of PM scheduling in two modes for more precise and better machine maintenance, as pieces must be replaced or be repaired. Because of the importance of this problem, we define multi-objective functions including makespan, PM cost, variance tardiness, and variance cost; we also consider multi-parallel series machines that perform multiple jobs on each machine and an aid, the analytic network process, to weight these objectives and their alternatives. PM scheduling is an NP-hard problem, so we use a dynamic genetic algorithm (GA) (the probability of mutation and crossover is changed through the main GA) to solve our algorithm and present another heuristic model (particle swarm optimization) algorithm against which to compare the GA’s answer. At the end, a numerical example shows that the presented method is very useful in implementing and maintaining machines and devices.     

互替机床提前/延期惩罚调度问题的启发式算法

对以作业提前或延期惩罚因素之和最小为目标函数的互替机床调度问题进行了描述，提出和阐述了一种四段式启发式算法，并通过大量不同规模的问题仿真对该算法进行了评价分析，结果表明该算法可行、有效。