可重构制造系统重构算法的实例研究

可重构制造系统（RcMS)的模型及其重构算法是可重构制造系统的形式化表示，是研究RcMS优化的理论基础，基于赋时可重构Petri网的可重构制造系统模型（TRPN－RcMS)及其重构算法是RcMS的形式化表示方法之一。文中以实例对RcMS重构算法进行说明和论证。首先假设一个RcMS的实例系统；然后，根据RcMS的当前系统组成及其生成计划生成RcMS当前系统的TRPN－RcMS模型；最后，根据RcMS重构后的系统组成及其生产计划的变化情况，利用RcMS的TRPN－RcMS模型的重构算法，生成重构后的系统的TRPN－RcMS模型。     

利用决策支持系统分析可重构制造系统中的问题

可重构制造系统是面向新世纪的先进制造模式 .本文提出利用决策支持系统解决可重构制造系统所面临的重构决策问题 ,在系统决策、加工单元布局、以及可重构产品的生产计划问题上采用智能化算法 ,可以得到满意的结果     

可重构制造系统的Petri网建模和分析方法

提出一种针对可重构制造系统的Petri网建模和分析方法．根据生产流程图可以得出制造系统的基本网模型,扩展基本网模型即可得到系统的Petri网模型．当生产任务发生改变并建立新的生产流程图时,可直接从原来的基本网模型构造出新构形的基本网模型．此外给出了系统重构代价的评价方法．仿真研究验证了该方法的有效性。     

基于RMC的可重构制造系统设备布局优化研究*

为实现多品种变批量生产制造系统阵列式布局的动态重构,提出一种新的设备布局优化方法.建立了可重构制造系统(RMS)设备优化选择数学模型,设计了基于蚁群优化和阶序聚类算法的可重构制造单元(RMC)动态重构算法.以交货期内最小成本为目标,引入系统复杂度和系统响应度从系统能力角度完善了实现RMS重构的约束条件,最终完成了可重构制造系统设备布局优化.最后通过布局实例仿真验证该方法的可行性和有效性.     

可重构制造系统故障诊断多Agent自学习模型

提出了一种基于多Agent的可重构制造系统故障诊断自学习模型,模型为3层结构,从下向上分别为监测Agent层、诊断Agent层和管理Agent层,详细分析了模型的故障诊断自学习机理,自学习性是基于数据挖掘和基于粗糙集的.示例表明可重构制造系统重构前后,均能有效完成故障诊断.     

基于Agent的可重构装配线制造执行系统

针对制造执行系统存在的问题,根据可重构制造模式的理念,设计基于CORBA和多Agent的可重构装配线制造执行系统的体系结构,实现制造执行系统的可重构性和可集成性,构建系统的IDEF0功能模型,给出Agent结构模型及装配资源聚类方法。实际应用表明该系统具有良好的实用性,能满足企业需求。     

基于多Agent的可重构制造系统集成模型

应用控制、管理和维护一体化的自动化技术,建立了基于多Agent的可重构制造系统RMS（Reconfigurable Manufacturing System）集成模型。该模型集成了基于多Agent 的RMS重构模型、控制模型和故障诊断模型,将RMS的控制、管理和维护联系起来,并给出了该模型的UML（Unified Modeling Langurage）活动图,最后举例验证了模型的可行性。     

可重构制造系统可重构逻辑控制器设计与实现

针对可重构制造系统的逻辑控制问题,提出一种可重构逻辑控制器的解决方案．该逻辑控制器具有递阶分布式的控制体系结构,并根据模块化的设计思想设计成多个分离的功能模块．然后给出基于CORBA组件模型(CCM)的可重构逻辑控制器软件的开发过程．由递阶分布式体系、模块化设计和软件组件开发技术实现的可重构逻辑控制器具有快速动态重构的能力,能满足可重构制造系统逻辑控制的要求．     

可重构制造系统的模糊综合评价研究及其PB实现

针对目前可重构制造系统评价体系的不足,根据可重构制造系统的特点和要求,以某研究所机械加工车间为研究对象,从工程应用的实际出发,构建了一个较为完备的可重构制造系统的评价体系模型,并通过层次分析法对可重构制造系统的各级指标进行量化,在此基础上,运用模糊综合评价方法对系统进行综合评价打分,从而达到对系统的可重构性进行控制的目的。并借助PowerBuilder软件进行评价界面的设计,解决了复杂的计算问题,提高评价的准确度与时效性。     

制造系统的可重构布局设计

研究重构系统的新布局具有重要的理论与实际意义 .本文讨论了制造系统可重构布局研究的目标、对象、基础及研究范畴 ,提出了利用模拟退火方法寻找重构布局最优解的搜索算法 .仿真实验验证了算法的有效性     

Automatic Reconfiguration of Petri Net Controllers for Reconfigurable Manufacturing Systems With an Improved Net Rewriting System-Based Approach

The advent of reconfigurable manufacturing systems (RMSs) has given rise to a challenging problem, i.e., how to reconfigure rapidly and validly a RMS supervisory controller in response to frequent changes in the manufacturing system configuration driven by fluctuating market. This paper presents an improved net rewriting system (INRS)-based method for automatic reconfiguration of Petri net (PN) supervisory controllers for RMS. We begin with presenting the INRS which overcomes the limitations of the net rewriting system and can dynamically change the structure of a PN without damaging its important behavioral properties. Based on INRS, a method for design reconfigurable PN controllers of RMS is introduced. Subsequently, we presented an INRS-based method for rapidly automatic reconfiguration of this class of PN controllers. In the reconfiguration method, changes in a RMS configuration can be formalized and act on an existing controller to make it reconfigure rapidly into a new one. Noticeably, no matter the design or reconfiguration, the expected behavioral properties of the resultant PN controllers are guaranteed. Thus, efforts for verification of the results can be avoided naturally. We also illustrate the reconfiguration of a PN controller for a reconfigurable manufacturing cell.     

Reconfiguration point decision method based on dynamic complexity for reconfigurable manufacturing system (RMS)

To address the problem of how to identify the best time to implement reconfiguration for the reconfigurable manufacturing system (RMS), a dynamic complexity-based RMS reconfiguration point decision method is proposed. This method first identifies factors that affect RMS dynamic complexity (including both positive and negative complexity) at the machine tool and manufacturing cell levels. Next, based on information entropy theory, a quantitative model for RMS dynamic complexity is created, which is solved via state probability analysis for processing capability and the processing function. This model is combined with cusp catastrophe theory to establish an RMS reconfiguration decision model. Both positive and negative complexity are control variables for cusp catastrophe. Cusp catastrophe’s state condition is used to identify RMS state catastrophe at the final stage of production. This catastrophe point is the RMS reconfiguration point. Finally, the case study result shows that this method can effectively identify the RMS state catastrophe moment so that system reconfiguration is implemented promptly to improve RMS’s responsiveness to the market.     

可重构资源管理及硬件任务布局的算法研究

可重构系统具有微处理器的灵活性和接近于ASIC的计算速度,可重构硬件的动态部分重构能力能够实现计算和重构操作的重叠,使系统能够动态地改变运行任务,可重构资源管理和硬件任务布局方法是提高可重构系统性能的关键.提出了基于任务上边界计算最大空闲矩形的算法(TT-KAMER),能够有效地管理系统的空闲可重构资源;在此基础上使用FF和启发式BF算法进行硬件任务的布局.实验表明,算法能够有效地实现在线资源分配与任务布局,获得较高的资源利用率.     

可重构制造系统监督控制器的自动重构

提出了基于改进的网重写系统(Improved net rewriting system, INRS)的可重构制造系统(Reconfigurable manufacturing systems, RMS) Petri网监督控制器的自动重构方法, 以快速适应由市场需求变化所引起的制造系统构形的频繁变化. INRS解决了网重写系统存在的问题, 可动态调整给定Petri网模型的结构而不改变其行为属性. 以集合和图的组合形式定义了RMS的构形, 并提出了基于INRS的一类模块化、可重构的Petri网控制器的设计方法. 针对这类Petri网控制器, 提出了基于INRS的自动重构方法. 方法可将RMS构形的变化转变为INRS的图重写规则, 并作用于当前Petri网控制器, 使其快速、自动地重构为所求的新控制器. 所提出的Petri网控制器的设计与重构方法, 均从理论上保证了结果的正确性, 免校验. 仿真研究验证了方法的有效性.     

基于遗传算法的可重构系统软硬件划分

在考虑动态部分重构及重构延时等特征的基础上,采用遗传算法及其与爬山算法的融合实现可重构系统软硬件任务的划分,并采用动态优先级调度算法进行划分结果的评价。实验表明,在可重构系统的资源约束等条件下,算法能够有效地实现应用任务图到可重构系统的时空映射。     

基于免疫重构的阴性选择算法

针对阴性选择算法在解决实际问题中,易误判及自修复能力差的弱点,该文基于生物免疫系统内部学习优化机制以及工业领域中的可重构系统,提出了一种基于免疫重构的阴性选择算法.新算法将可重构系统的思想融入到阴性选择算法中,提出了重构串、重构模型与重构操作的概念与实现方法,以保证系统发生意外的时候能够及时恢复、重组.将算法应用于一个Web系统进行仿真实验,结果表明该算法是有效的.     

Rapid design and reconfiguration of Petri net models for reconfigurable manufacturing cells with improved net rewriting systems and activity diagrams

To respond rapidly to the highly volatile market, the emerging reconfigurable manufacturing systems (RMS) have brought forward two challenging issues, namely, how to build rapid a formal model of an initial manufacturing configuration and how to yield the goal model from the existing one along with manufacturing configuration changes (reconfiguration). As for the issues, we present in this paper a method for rapid design of Petri net (PN) formalized models of RMS, intended for supervisory control and logic control of RMS, as well as a method for automated reconfiguration of the models. Firstly, we present an improved net rewriting system (INRS) for dynamically operating net transformation, unlike its predecessor-net rewriting system, where the initial behavioral properties of the underlying PN rewritten can be preserved during the transformation. Subsequently, the paper proposes the three-phase method for rapid design of initial full PN models of reconfigurable manufacturing cells (RMCs). In this method, activity diagrams of Unified Modeling Languages version 2 (UML 2) are used to describe manufacturing configurations, firstly; then the sub-activity diagrams are transformed into PN sub-models; finally, the PN sub-models are automated synthesized into a full model by the approach of INRS. Further, we present a model reconfiguration method for this class of PN models. The method compares changes in activity diagrams of the existing and goal manufacturing configurations and converts them into net rewriting rules of INRS. By applying the rules obtained, the existing PN model can reconfigure into a new one for the goal manufacturing configuration. No matter the design method or the reconfiguration method, the behavioral properties of the obtained PN models, e.g., liveness, boundedness, or reversibility, can be guaranteed and thereby the efforts of verification can be avoided. Finally, rapid design of a PN model of a reconfigurable manufacturing cell, as well as its automated reconfiguration, is illustrated with the help of an example. The result indicates the validity of the methods.     

诊断多智能体重构过程的系统描

在引入可重构故障诊断系统及诊断智能体单元组概念的基础上，提出了可重构故障诊断系统的体系结构和应用系统模型，探讨了被诊断问题的特征以及被诊断问题之间的相似性，构建了基于决策理论的多智能体诊断系统的评价方法．在分析诊断智能体重构过程的基础上，研究了可重构故障诊断系统的重构类型、重构层次和重构步骤．借鉴生物免疫中的调节理论，提出了一种重建故障诊断智能体的算法．最后分别就智能体单元组（AFDS）的物理重构和逻辑重构两种形式进行了举例说明．