INTEGRATION OF GENERIC SHOP FLOOR CONTROL

Based on the discussion of functions, components, control structure and network computing architecture of generic shop floor control (GSFC) in computer integrated manufacturing system (CIMS), the internal integration between components of GSFC and the external integration between GSFC and the other components of CIMS are studied. The integration models on the aspects of function, information, processing and interface are put forward. The integration models and the methods are implemented and applied in CIMS projects successfully.

An Improved PAC Architecture for Shop Floor Information System in Digital Manufacturing

In order to overcome the weakness of the former production activity control （PAC） architecture, this paper presents an improved PAC architecture. The BIT-PAC architecture has been applied with the use of information technology. It facilitates the information flow from Shop Floor to other areas of the organization and makes the organizations more integrated and productive. This architecture also facilitates the expansion of the Shop Floor functions without disturbing the basic infrastructure. Effectiveness of the BIT-PAC architecture was checked by developing and running application software on a network PC which has supported the dynamic flow of information from Sub-Producers and Sub-Movers to other areas of organization.

Model of Production Planning and Scheduling in Shop Floor Based on Collaboration and Competition

The planning and scheduling in real shop floor is actually achieved by coordination between different persons. In this process, cooperation is mainstream, but competition also exists, for example, the competition between different groups, operators with the same skill, etc. In multi-agent based shop floor management and control system, this competition and cooperation relation must be embodied. The general process of shop floor production planning and scheduling is studied, and a colored Petri-net model for the competition and cooperation process of three main agents in such system to achieve shop floor production planning and scheduling is presented. The evaluating method of bids in bidding process that especially embodies the competition relationship is also presented. This colored Petri-net model gives a clear illustration of this complex coordination process to system designers, effectively promotes the cooperative development.

MODELING AND IMPLEMENTATION OF A DISTRIBUTED SHOP FLOOR MANAGEMENT AND CONTROL SYSTEM

Adopting distributed control architecture is the important developmentdirection for shop floor management and control system, is also the requirement of making it agile,intelligent and concurrent. Some key problems in achieving distributed control architecture areresearched. An activity model of shop floor is presented as the requirement definition of theprototype system. The multi-agent based software architecture is constructed. How the core part inshop floor management and control system, production plan and scheduling is achieved. Thecooperation of different agents is illustrated. Finally, the implementation of the prototype systemis narrated.

An Improved Production Activity Control Architecture for Shop Floor Control

This paper presents a further improved Production Activity Control Architecture to deal with the complexity of information by creating Sub-Producers and Sub-Movers which will not only give a better control at workstation level but also reduce load on the Dispatcher. It also makes an analysis of the basic and improved PAC (Production Activity Control) Architecture in the Control System for Integrated Manufacturing. The PAC Architecture and the improvement will further enhance the flexibility and adaptability of the architecture in the ever changing environment of the Shop Floor Control (SFC) Systems.

A PSL Ontology-based Shop Floor Dynamical Scheduler Design

Due to the complex,uncertainty and dynamics in the modern manufacturing environment,a flexible and robust shop floor scheduler is essential to achieve the production goals.A design framework of a shop floor dynamical scheduler is presented in this paper.The workflow and function modules of the scheduler are discussed in detail.A multi-step adaptive scheduling strategy and a process specification language,which is an ontology-based representation of process plan,are utilized in the proposed scheduler.The scheduler acquires the dispatching rule from the knowledge base and uses the build-in on-line simulator to evaluate the obtained rule.These technologies enable the scheduler to improve its fine-tune ability and effectively transfer process information into other heterogeneous information systems in a shop floor.The effectiveness of the suggested structure will be demonstrated via its application in the scheduling system of a manufacturing enterprise.

APPLICATION OF IMPROVED PRODUCTION ACTIVITY CONTROL ARCHITECTURE FOR SHOP FLOOR INFORMATION SYSTEM IN DIGITAL MANUFACTURING

Shop floor control （SFC） is responsible for the coordination and control of the manufacturing physical and information flow within the shop floor in the manufacturing system. Weaknesses of the production activity control （PAC） architecture of the shop floor are addressed by the Maglica＇s new system architecture. This architecture gives rise to unlimited number of movers and producers thus evolving more complex but decentralized architecture. Beijing Institute of Technology - production activity control （BIT-PAC） architecture introduces an idea of sub-producars and sub-movers thus reducing the complexity of the architecture. All the equipments including sub-producars and sub-movers are considered to be passive in the proposed shop floor information system. The dissemination of information from sub-producers and sub-movers is done manually through a PC. Proposed BIT-PAC SFC architecture facilitates the information flow from shop floor to the other area of the organization. Effective use of interact information services （IIS） and SQL2000 is done along with the ASP.NET technology to implement the application logic. Applicability of the software based on BIT-PAC architecture is checked by running application software on a network PC that supports the dynamic flow of information from sub-producers and sub-movers to the other parts of the organization. Use of software is also shown at the end for BIT training workshop thus supporting the use of SFC architecture for similar kind of environments.

Proposal for a Project of Integration of IT Tools With Focus on Planning and Control of Operations on the Shop Floor

Manufacturing system, with high level of complexity and with a mix of semi-repetitive and repetitive products, to become productive, should seek the standardization of products and processes to obtain the optimization of use of production resources. However, it is necessary to measure the productivity, so that the system of measurement and control of manufacturing processes are an element critical as to ensure greater visibility of the flow＇s restrictions, minimized when detected properly. In this case, the automation of factory＇s measurement process can effectively contribute to ensuring the effectiveness of the function control of a manufacturing system. It is important to consider that the automation of the system of measurement and control of manufacturing processes, of complex environment, is heavily dependent of IT tools applied directly in the interface computational between the operation systems and the corporate systems. This heavy reliance, if exploited technically properly, allows that automation of the system of measurement and control of production makes the access to time real of availability of manufacturing process＇s data, such as processing time and setup time that it can export to a specialist software in programming production, for example, feasible. In this paper, the automation of the system of measurement and control of production is approached, in order to identify the main possibilities of the design of an information system capable to integrate the flow of information in an environment internal on manufacturing organizations, with emphasis in the digital manufacturing paradigm.

Performance Improvement in Manufacturing Shop Floor Operations of Developing Countries Based on Three Characteristics of Information Flow

The management of information flow for production improvement has always been a target in the research. In this paper, the focus is on the analysis model of the characteristics of information flow in shop floor operations based on the influence that dimension (support or medium), direction and the quality information flow have on the value of information flow using machine learning classification algorithms. The obtained results of classification algorithms used to analyze the value of information flow are Decision Trees (DT) and Random Forest (RF) with a score of 0.99% and the mean absolute error of 0.005. The results also show that the management of information flow using DT or RF shows that, the dimension of information such as digital information has the greatest value of information flow in shop floor operations when the shop floor is totally digitalized. Direction of information flow does not have any great influence on shop floor operations processes when the operations processes are digitalized or done by operators as machines.

A HOLONIC FRAMEWORK OF RECONFIGURABLE SHOP FLOOR CONTROL SYSTEMS IN AUTOMATED MANUFACTURING SYSTEMS

The agility and the flexibility of the current shop floor control systems have been limited so far, owing to the lack of structural flexibility and agility in its control software layer. Most of them are based on traditional hierarchical architecture and the top down approach and depend structurally on their specific configuration and job scheduling. Not only can they hardly satisfactorily adapt to these increasing changes and disturbances, but also make the redevelopment and maintenance of shop floor control system (SFCS) to need high cost and much time. And SFCS based on the heterarchical architecture don′t provide a predictable and high performance system, especially not in the heterogeneous environments, where the resources are scarce and the current decisions have serious repercussions on the future performances. For this reason, the heterarchical control is hardly applied in industry. Obviously, it is necessary to develop a new structural framework of reconfigurable SFCS to improve their agility, flexibility and maintainability. This paper presents a holonic framework of reconfigurable SFCS based on holonic manufacturing concepts. The framework is composed of resource holons, product holons and other staff holons. The model of each holon and the co operative mechanisms of holons are described. To verify the proposed approach experimentally, a prototype reconfigurable SFCS for a flexible manufacturing shop floor producing discrete parts is implemented.

数字孪生车间制造运营管理平台

数字孪生使能的制造信息物理融合是数字经济时代制造业发展的主流方向,数字孪生车间作为制造基础单元,能够基于可信数据和模型的融合决策实现面向多场景适配的制造运营管理能力提升。针对制造运营管理精益化、柔性化、智能化发展需求,以及制造运营管理能力与物理需求难适配、制造运营管理模式与物理状态难适用、制造运营管理策略与物理场景难适应的问题,提出了数字孪生车间制造运营管理平台,阐述了其系统组成、特点和关键技术。在此基础上,随着制造运营管理向着包含物理资源和虚拟资源的跨组织、跨企业智能化协同运营延伸,提出了跨组织/跨企业的“组织管理规范洞察决策改善”协同运营管理策略,为支撑车间内、企业内、跨企业等制造运营管理理论研究和应用落地提供系列理论和方法参考。

基于分解协调机制的多层单元流水式车间布局优化方法

针对多层单元流水式车间,考虑其具有随机性因素的特点,研究了该车间的布局优化方法。首先,以最小化总运输成本为目标函数,建立具有产出率和生产周期等约束的优化模型;其次,根据系统性能指标没有封闭数学表达式的问题特点,提出一种基于分解协调的优化算法,同时嵌入排队网模型进行迭代求解;最后,设计了一系列算例实验验证所提方法的有效性和优越性,并将其代入实际生产车间进行案例分析,为该类生产车间的布局规划问题提供重要的决策依据。

基于改进强化学习的准时化物料搬运系统实时调度方法

准时高效的物料搬运系统保证了装配制造的持续稳定运行,为动态应对装配线状态变化,有效平衡混流装配的生产效率与能耗,本文提出了基于Q学习算法的强化学习调度模型,对其系统状态、动作策略、报酬函数进行设计,并引入神经网络对Q值函数进行泛化和逼近,改进策略选择机制,形成基于双参数贪婪策略的强化学习动态调度方法。仿真实验结果表明,这种强化学习调度相比其他调度方法,物料搬运调度的优化效果更好,能在保证物料准时运送到装配线,实现最大产量的同时,有效减少搬运距离。

Energy-efficient Approach to Minimizing the Energy Consumption in An Extended Job-shop Scheduling Problem

The traditional production planning and scheduling problems consider performance indicators like time, cost and quality as optimization objectives in manufacturing processes. However, environmentally-friendly factors like energy consumption of production have not been completely taken into consideration. Against this background, this paper addresses an approach to modify a given schedule generated by a production plarming and scheduling system in a job shop floor, where machine tools can work at different cutting speeds. It can adjust the cutting speeds of the operations while keeping the original assignment and processing sequence of operations of each job fixed in order to obtain energy savings. First, the proposed approach, based on a mixed integer programming mathematical model, changes the total idle time of the given schedule to minimize energy consumption in the job shop floor while accepting the optimal solution of the scheduling objective, makespan. Then, a genetic-simulated annealing algorithm is used to explore the optimal solution due to the fact that the problem is strongly NP-hard. Finally, the effectiveness of the approach is performed small- and large-size instances, respectively. The experimental results show that the approach can save 5%-10% of the average energy consumption while accepting the optimal solution of the makespan in small-size instances. In addition, the average maximum energy saving ratio can reach to 13%. And it can save approximately 1%-4% of the average energy consumption and approximately 2.4% of the average maximum energy while accepting the near-optimal solution of the makespan in large-size instances. The proposed research provides an interesting point to explore an energy-aware schedule optimization for a traditional production planning and scheduling problem.

基于量子计算和威布尔分布的混合CHIO算法求解JSP问题

针对冠状病毒群免疫优化算法(coronavirus herd immunity optimizer,CHIO)在解决优化问题时存在易陷入局部最优解、收敛速度慢和收敛精度差等问题,文章提出一种量子混合CHIO算法(quantum hybrid coronavirus herd immunity optimizer,QCHIO)。首先,引入量子计算的思想,通过量子相关性实现全局搜索和快速收敛的目标,能够有效避免算法陷入局部最优解的问题。其次,采用威布尔分布算子的大步长和小步长来增加算法的多样性,使算法能够更好地探索搜索空间,增强了算法的全局开发能力。此外,还引入β-登山算子通过搜索当前最优解的邻域,尝试找到更优的解,从而增加了算法的搜索宽度,改善了解的质量。多邻域搜索则通过搜索全局最优解的多个邻域来增加了算法的收敛精度。为验证其性能,将QCHIO应用到10种标准算例中与其他几种改进算法进行了对比分析,并通过显著性检验证明了QCHIO的优越性。最后将QCHIO应用到某发动机生产调度实例上,进一步证明了QCHIO的可行性和优越性。

基于BOM生产的车间排产多约束条件问题研究

随着生产制造业的转型升级,企业对于物料的生产不再是预先生产,更多是根据实时订单生产,针对这一情况,就需要考虑物料的优先生产问题,该研究将“产品物料优于产品先加工生产”作为约束条件之一,在传统遗传算法的基础上,提出对染色体进行二次编码,设计基于BOM(物料清单)的变异算子。从订单完工时间、设备停滞运行时间、两者比值3个维度分析不同的可行方案,给出最优排产计划。实验表明,该改进遗传算法在满足“产品物料优于产品先加工生产”在内的所有约束条件的情况下从3个维度分析后产生了最优排产计划。该算法具有普适性,有较强的参考价值。

隐性扰动下智能车间资源重调度决策方法研究

针对智能制造车间资源重调度快速响应的需求,以及隐性扰动难以测量捕捉的特点,提出了车间资源监测及重调度决策方法。首先利用支持向量机良好的连续监测性能,建立了资源异常状态监测模型;其次通过结合lasso回归算法和K近邻值分类算法提高SVM模型的预测值精准度,利用数据替代手段构建容错机制,保证系统异常时的短暂平稳运行;然后设计了车间资源重调度方式,通过历史案例数据训练分类器用于重调度方案抉择,指导智能制造车间在隐性扰动情况下的高效生产;最后,以实际车间隐性扰动为例,验证了所提的重调度决策方法的有效性。

基于双深度Q学习网络的面向设备负荷稳定的智能车间调度方法

在车间管理中,设备负荷是一个关键性能指标,负荷稳定直接影响了生产效率与生产成本,但目前鲜有研究关注如何实现设备负荷稳定的问题。为此,提出一种面向设备负荷稳定的智能车间调度方法。该方法通过一个含有深度神经网络调度模型的调度智能体,分析车间生产状态与设备负荷间的相关性,及时输出满足期望目标的调度方案。针对深度神经网络调度模型,设计了一个基于双深度Q学习网络(DDQN)的深度神经网络调度模型训练器,其利用奖惩学习免监督地形成调度样本,借此对深度神经网络调度模型进行网络参数更新,实现模型自学习。所提方法在MiniFab半导体生产车间模型中进行了验证,证明了所提调度方法能实现对智能车间设备负荷的控制,从而保证车间整体设备负荷的稳定性。

基于Q学习的智能车间自适应调度方法

为降低智能车间中动态不确定因素对生产性能的影响,提出一种基于Q学习的智能车间自适应调度方法。该方法设计基于强化学习的智能车间自适应调度框架,采用Q学习算法,通过智能体—环境交互试错机制,自主训练调度模型,并根据生产车间环境变化动态更新调度模型,以支持能够指导车间运行的最优决策轨迹的生成。所提方法在MiniFab半导体生产线模型上进行了验证,结果证明该方法能够有效应对智能车间生产环境变化,在生产全过程中能对调度决策进行实时调整,优化车间综合性能指标,同时显著降低时间与人力成本。

运输资源共享的多层单元流水式车间排队网建模与分析

针对多层单元流水式车间,着重考虑其运输资源共享的特点,研究该车间的排队网建模方法。首先,定义不同任务的可能场景及其发生概率,采用两矩近似法描述任务节点的服务时间;其次,提出一种混合排队网模型处理车间中的同步性约束问题,并根据这些约束建立模型中各节点的关系,考虑一般分布和有限缓存区,通过迭代近似求解系统的性能指标;最后,设计了一系列的算例实验并与仿真结果对比,验证了假设模型的有效性和求解方法的精确性,并分析运输设备和缓存区对系统性能的影响,为该类生产车间的资源配置和布局规划等问题提供重要的理论支撑。