考虑多目标的柔性工艺与调度集成优化算法

针对工艺规划与车间调度集成优化问题,在考虑零件的加工工序柔性、工序次序柔性及加工机器柔性的基础上,以最大完工时间、总加工成本和总拖期时间为优化目标,对多目标柔性工艺与车间调度集成问题建模,提出一种基于改进人工蜂群算法的多目标柔性工艺与车间调度集成优化策略,并提出邻域变异操作以及全局交叉操作,对种群进行更新。引入Pareto方法,通过对适应度评价、贪婪准则、Pareto最优解集构造和保存以及解得多样性维护等方面进行改进,设计了一种基于Pareto方法的多目标人工蜂群算法。最后,通过采用基本人工蜂群算法及改进人工蜂群算法对六个工件、五台机床的柔性工艺与车间调度集成问题进行优化,验证了改进算法的有效性。     

基于多Agent协商的柔性车间调度系统

针对柔性作业车间调度的特点,提出了一种基于多agent协商的柔性作业车间调度系统。系统由工件agent,机器agent和工序agent组成。Agent之间通过相互发送消息和响应消息进行交互,并且通过消息相应函数按照各agent局部的信息、同时兼顾系统的性能进行决策。工件agent通过招标的方式,选择合适的机器完成加工任务,机器agent通过竞争来获得工件的加工权。最后用Java语言在Eclipse平台上进行程序设计,对柔性作业车间调度的平均滞后问题进行仿真实验,并与传统的分派规则比较,结果显示所提方法的优越性。     

基于人工蜂群算法的柔性工艺与车间调度集成优化

为实现柔性工艺与车间调度集成优化,在考虑工件特征的加工工艺、次序及加工机器的柔性基础上,以最小化最大完工时间为优化目标,提出一种基于交叉变异的人工蜂群算法。该算法针对柔性工艺与车间调度集成问题的离散性特征,对工艺路线进行序列编码,工件调度采用基于工序的编码方式。通过工艺种群与调度种群的交叉变异操作,分别使采蜜蜂及观察蜂进行局部寻优,侦查蜂进行全局寻优,以此提高算法性能。在此基础上用两部分测试实例分别验证了集成研究的必要性及改进算法的有效性。     

基于非线性工艺规划思想的车间动态调度系统

描述了一种基于非线性工艺规划思想的车间动态调度系统的集成框架结构，其核心模块为计划调度模块，该模块调用由遗传算法和启发式调度相结合的调度算法生成动态调度方案。其中所提出的遗传编码的设计充分考虑工艺路径的柔性，并根据此编码提出调度方案的构造方法，同时相应地改进了遗传操作，从而实现了调度的全局最优性和可行性。     

面向柔性作业分布式车间的分层调度模型研究

针对多车间分布式制造系统调度优化问题,结合车间实际生产情况,提出一种基于目标级联法和遗传算法的层次调度模型。模型将生产调度过程划分为生产计划层、车间调度层和零件规划层,并将整体时间最短的优化目标划分到各个层次,通过层层优化达到时间最优后反馈至上层,以实现整体调度时间最优。以3个制造车间协调调度问题为例,验证了该模型在零件分配和零件的工艺路线选择上的合理性和有效性。     

基于多Agent柔性制造单元的生产调度研究

针对基于制造单元的作业车间的生产调度问题进行了研究,结合多代理的智能性、灵活性和遗传算法的智能优化能力,建立基于多智能体的柔性制造单元的作业车间的调度系统模型.然后,提出了集成多智能体和遗传算法的动态调度策略和调度协商机制;最后,应用此方法完成了常规调度和异常调度的仿真算例.结果表明所开发系统可以解决基于加工单元的制造...     

基于需求优先的多目标柔性车间调度研究

为满足按时提交客户货物的要求,需要优化企业的生产调度,现实的生产调度问题是传统车间调度问题的扩充,具有多目标、柔性等特性。针对柔性作业车间调度的需要,提出了在精益制造下的基于需求优先的多目标柔性车间调度算法。该算法以工件提前／拖期惩罚代价最小,调度最小生产周期为目标,基于规则的改进启发式调度,在调度过程中通过需求日期计算工件的优先级为每道工序分配合适的机器进行加工,可得到满意的较优解。与其他方法进行对比试验的结果表明,该算法在求解柔性作业车间调度问题是有效的。     

一种改进的蚁群算法在工艺规划与车间调度集成优化中的应用

改进标准蚁群算法的执行策略,可提高工艺规划和调度集成问题的求解 质量和效率。通过节点集、有向弧/无向弧集、AND/OR 关系,建立了基于AND/OR 图的工 艺规划和调度集成优化模型。提出一种求解工艺规划与车间调度集成问题的改进蚁群优化算 法,采用了信息素动态更新策略避免收敛过慢和局部收敛,利用多目标优化策略提高求解质 量。仿真结果证明了该算法的有效性。     

基于遗传算法的航空零件加工柔性车间调度研究

车间调度问题计算复杂,约束条件度,一般算法难以实现全局搜索,算法比较容易陷入局部最优。通过对柔性车间调度问题的深入研究,采用遗传算法进行柔性车间调度,使一个个体可以表达全部零件加工顺序,并用适应度函数平价个体好坏。种群通过选择算子、交叉算子和变异算子不断进化,最终得到最优的柔性车间调度方法。通过仿真试验表明,该算法能够有效地进行车间调度。     

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。     

A Case Study on Integrated Production Planning and Scheduling in a Three-Stage Manufacturing System

This paper presents an integrated optimization model of production planning and scheduling for a three-stage manufacturing system, which is composed of a forward chain of three kinds of workshops: a job shop, a parallel flow shop consisting of parallel production lines, and a single machine shop. As the products at the second stage are assembled from the parts produced in its upstream workshop, a complicated production process is involved. On the basis of the analysis of the batch production, a dynamic batch splitting and amalgamating algorithm is proposed. Then, a heuristic algorithm based on a genetic algorithm (known as the integrated optimization algorithm) is proposed for solving the problem. Note to Practitioners-This paper presents a method for integrated production planning and scheduling in a three-stage manufacturing system consisting of a forward chain of three kinds of workshops, which is common in such enterprises as producers of automobiles and household electric appliances, as in the case of an autobody plant usually with the stamping workshop, the welding and assembling workshop, and the painting workshop. Herein, the production planning and scheduling problems are simultaneously addressed in the way that a feasible production plan can be obtained and the inventory reduced. A batch splitting and amalgamating algorithm is proposed for balancing the production time of the production lines. And a case study of the integrated planning and scheduling problem in a real autobody plant verifies the effectiveness of our method     

基于模型重构的生产计划优化系统设计与开发

敏捷制造是一种可以快速响应客户需求的制造模式,但现有敏捷制造模式下的生产计划优化系统存在针对性弱、生产模型单一等问题。文中通过提取组成生产目标规划模型的各种因子,对敏捷制造模式下满足不同生产目标和生产条件的生产模型进行了重构,并采用卡马卡及其关联预测算法进行了模型求解,从而大大提高了车间生产计划优化系统的灵活性和适用性,为生产效率的提高提供了一条更加科学有效的途径。     

A Petri net model for integrated process and job shop production planning

In this paper a new Petri net class (Chameleon systems) for modeling and analyzing joined process planning and job shop production planning is proposed based on the multi-level Petri net model introduced by Valk (1995). Chameleon system is a High-level Petri net extended with time that consists of a High-level Petri net modeling the workshop which is called system net and ordinary Petri net modeling the jobs and setups which act as tokens of the system net and therefore are called token nets. Process plans of the jobs are partial orders of operation groups and alternatives. The uncertainty of operation duration is captured by a new time model. The advantages of this new modeling approach are the following: (i) a modular construction of the joined process and job shop planning is allowed, (ii) classical known Petri net analysis methods can be applied and (iii) the analysis can be done at different levels of abstraction.     

批量生产企业的车间计划调度算法研究与应用

分析了批量生产企业车间计划调度问题的资源、任务和活动,讨论了车间计划调度的约束及其实现,给出了基于复合规则的计划调度算法,在西安航空发动机公司32车间的计划管理系统中得到了应用。     

An expert system for process planning

Fully integrated manufacturing systems, the ultimate goal of today's production engineering research, cannot be realized without automated process planning systems. A knowledge-based expert system incorporating heuristic algorithms as well as analytical and empirical models and which can thus provide a partial replacement for human expertise is the only key to automated process planning.This paper presents an attempt to develop an expert system for automatic process planning using artificial intelligence. In this system, a structural database was incorporated with a knowledgebase for the execution of rules. The programs developed are user-friendly and highly interactive. They allow dialog between the system and the user and require only that the operator answer questions in a familiar workshop language.     

基于增强现实的制造数据记录系统设计与实现

针对传统制造业数字化水平不高的问题,提出了基于增强现实的生产数据记录系统;分析了车间数字化转型的软件架构需求,应用了 RESTful风格的前后端分离软件架构,将系统数据库和前端应用解耦,实现了基于HTTP协议的网络通讯;通过包含BLSTM的深度学习算法,对设备仪表图像进行检测识别,实现了制造过程关键生产数据智能化获取;通过对生产流程的分析和多维信息之间的联系,建立了基于增强现实的引导式生产数据记录应用;生产实践证明,系统能够实现生产任务的引导和过程数据的记录,工人工作效率提高,系统响应迅速,有利于车间生产过程的数字化转型.     

多智能体的车间染缸集群监控系统

针对染色车间染缸集群集中监控的问题,通过现场工业总线技术和以太网技术,建立了染缸过程控制、数据通信和集中临控三位一体的系统构架,并提出了一种多智能体的染缸集群监控方法.该方法实现了染缸集群的生产计划、作业任务、生产过程、染色工艺温度、染缸能源消耗等的实时监控.某大型印染车间的染缸集群监控系统应用结果表明,该系统取得了满意的实际效果.     

中药制造企业生产计划管理研究

中药生产属于流程行业是一种典型的批流程生产形式,它对批次的跟踪有着特别的要求。整合整个中药生产企业的制造资源对于实现中药生产企业的现代化管理是必不可少的。本文旨在通过对中药制造企业的生产过程分析,结合制造资源计划（Manufacturing Resource Planning,MRPII）原理与方法,对中药制药业的生产计划如何分解,以及如何形成作业计划进行分析和研究,提出符合其行业特点的计算模型和方法,为中药制造企业自动生成生产作业计划提供依据。     

Significance of the dimensional view for visualizing relevant aspects of a production system in a co-operative planning process

Pictorial visualization is expected to facilitate communication between industrial professionals when planning working environments and production systems. This hypothesis was investigated by studying how 24 participants including managers, supervisors, machine operators, and occupational health and safety officials, judged three types of computer animated visualization varying in dimensional view (scale and scope of a production line): shop floor view/survey of shop floor; production unit view/semi-survey of production unit; and workplace view/close-up of workplace, in relation to a set of planning issues. The participants participated in a controlled 2-day planning workshop, redesigning a fictitious manufacturing process by means of computer graphics, and then responded to a questionnaire. It can be concluded that shop floor view as well as production unit view are significant for survey planning issues, while all 3-dimensional views are significant for close-up planning issues. Analogously, all dimensional views are significant for technocentric planning issues, whereas only the workplace view is valuable for anthropocentric planning issues.