一个复杂工业过程控制MAS的设计与实现*

针对复杂工业过程控制的特点,建立了由控制Agent和决策支持Agent组成的分层递阶MAS(Multi Agent System);研究了基于Ice中间件的MAS开发方法,采用COM技术构建了具有二进制重用性的Agent组件,可实现对原有系统的重用;实现了MAS与专家系统、神经网络和模糊控制技术的融合;开发了新型干法水泥生产过程控制MAS,实现了水泥熟料煅烧过程的优化控制.     

Agent应用系统模板语言研究

为了有效地构造可靠、便于维护的多Agent系统（MAS），根据Grammarware的工程化思想与产生式编程技术，以基于黑板的单Agent结构为基础，提出一种Agent模板语言（APL）。通过一个企业业务过程管理（BPM）场景，说明在应用中如何使用APL描述Agent中针对具体应用的数据类型、知识源组件等，并由解析器将APL转换为JADE上的执行代码，进而实现多Agent系统的快速构建，实现MAS自顶向下的开发，从而提高系统开发的效率，增强系统的灵活性与可复用性。     

基于MAS的动态生产调度与控制及系统开发

提出基于MAS的面向敏捷制造的生产过程动态调度与控制的层次结构．1）以任务分解与分配层为中心，建立各层之间的协调工作及协同决策机制；2）引入协商式招／投标方法实现任务的分解与分配；3）采用能力匹配与动态调度相结合的方法实现任务分配与调度控制的有效集成；4）面向生产任务需求动态确定Agent粒度、组建MAS模型；5）适应制造系统状态变化的需要，进行任务的动态重构．讨论基于MAS的采用分级递阶和并行处理相结合的自治组织结构和运作模式，以及利用与组织结构相对应的层次黑板结构实现各Agent之间信息与数据共享．在支持生产过程动态调度与控制基础设施建设的基础上，结合奏川机床集团有限公司车间生产实际，研究开发了基于MAS的车间动态调度系统．     

MAS系统的问题求解能力分析

本文用状态空间搜索模型分析了多Agent系统(MAS)的问题求解能力，认为MAS系统中Agent之间知识的组合应用和对问题搜索方向的交互和决策是影响MAS系统问题求解能力的主要原因，在状态空间搜索模型下可以将Agent间知识的组合应用表达为不同Agent的搜索路径的组合，而Agent对搜索方向的判断是基于启发式信息做出的，从而为形式化分析MAS系统的性能建立了通用的模型．本文以A*算法为例探讨了可采纳算法下多Agent合作求解效果与Agent的知识和启发信息之间的关系，指出只有在一定条件下MAS系统才会获得更好的解题能力．本文还对非可采纳算法下MAS系统性能分析方法提出了初步看法．     

基于多Agent的特征提取模型研究

讨论了基于多Agent的特征提取模型的工作流程,主要研究了该多Agent模型的构建框架和模型中单个Agent的通用结构,借助JAFMAS这一面向MAS的基于Java的Agent框架实现了系统的建立,并最后给出了所举例子的实际运行结果。     

基于MAS的农业专家系统平台的研究

针对农业领域知识的分布性、易更新性、可重用性等特点,将MAS理论和技术应用到农业专家系统平台的开发中,提出了基于MAS的农业专家系统平台的结构,并讨论了实现该平台的一些关键技术.本系统平台的研究将加速Agent的实现与应用;同时会降低开发农业专家系统的成本,最终使农业专家系统更好地为农业、农村、农民服务.     

多 Agent 系统仿真平台

Agent和多Agent（MAS）系统目前已经逐步走向成熟化,且已经被广泛应用于系统建模、敏捷制造、网络监测等众多领域。为了能够便于Agent系统的开发,人们构建了各种Agent基础设施平台,例如已经广泛应用的JADE等。但是基于这些平台的应用大多具有平台依赖性,难于扩展和定制,无法支撑我们大规模的扩展使用。因此,将通过多Agent系统仿真平台Maze的设计与实现,着力构建出Agent系统各种基础设施的通用开发模式,包括Agent自治机制、通信机制、协作机制、协调机制等,从而便于用户在此基础上进行定制开发。另外,该平台的形象演示可以帮助初学者理解Agent的各种特性（如自治性、社会性、预动性等）,比较各种协作算法的优劣等,是一个很好的入门学习工具。     

基于MAS的农业专家系统平台的研究

针对农业领域知识的分布性、易更新性、可重用性等特点，将MAS理论和技术应用到农业专家系统平台的开发中，提出了基于MAS的农业专家系统平台的结构，并讨论了实现该平台的一些关键技术。本系统平台的研究将加速Agent的实现与应用；同时会降低开发农业专家系统的成本，最终使农业专家系统更好地为农业、农村、农民服务。     

基于多Agent的会议制订系统MSS/MAS

MSS/MAS系统是一种基于多Agent的分布式会议自动制订系统。它的协商方式是基于非终止主动合作计算模型,具有良好的理论背景。我们利用多Agent间的合作关系设计系统并实现高效的会议制定功能,因此MSS/MAS系统是一种具有较强应用价值的多Agent系统。     

MAS技术在ITS中的应用研究

多代理技术(MAS)是当今世界研究的热点。该文结合智能交通系统(ITS)具体应用,详细介绍了智能信息处理领域中MAS技术在ITS中的运用,提出了基于MAS技术的ITS系统构架,设计了智能Agent模块,实现了车辆检测Agent和参数Agent处理,具有较强的实用性。最后,讨论了MAS在ITS中的网络化设计。     

Construction method of shop-floor digital twin based on MBSE

Digital twin (DT) technology is essential for achieving the fusion of virtual-real cyber-physical systems. Academics and companies have made great strides in the theoretical research and case studies of constructing the shop-floor digital twin (SDT), which is the premise of applying DT technology on the shop floor. A shop floor is a large complex system that involves many elements including people, machines, materials, methods, and the environment and processes, such as the technical flow, business process, logistics, and control flow. However, most of the developed cases lack a hierarchical, structured and modularized implementation framework for the development of an SDT system, which leads to problems such as a low reuse rate of the system blocks, lack of scalability, and high upgrade and maintenance costs. In response to these issues, we propose a construction method of the DT for the shop floor based on model-based systems engineering from the perspective of the system. In this method, a comprehensive DT model for the shop floor is gradually constructed by using system modeling language, the modeling method “MagicGrid,” and the “V model” of systems engineering. The model includes four dimensions of the shop-floor requirements, structure, behavior, and parameters, as well as three stages (the problem domain, solution domain, and implementation domain), and connects nine steps of the “V model,” including the system requirements, system architecture, subsystem implementation, subsystem integration, and system verification. Then, based on an example of a real NC machining shop floor, subsystems including a visualization system, synchronization system, and simulation system, are discussed. Finally, the functions of the integrated systems are verified based on the requirements, including the real-time synchronization of “man, machine, material, and method” and the transient simulation in real time. The numerical indicators of the integrated system are verified, including the model completeness and synchronization timeliness.     

Applications of agent-based systems in intelligent manufacturing: An updated review

Agent technology has been recognized as a promising paradigm for next generation manufacturing systems. Researchers have attempted to apply agent technology to manufacturing enterprise integration, enterprise collaboration (including supply chain management and virtual enterprises), manufacturing process planning and scheduling, shop floor control, and to holonic manufacturing as an implementation methodology. This paper provides an update review on the recent achievements in these areas, and discusses some key issues in implementing agent-based manufacturing systems such as agent encapsulation, agent organization, agent coordination and negotiation, system dynamics, learning, optimization, security and privacy, tools and standards.     

一个基于多代理的并行工程设计系统模型研究

并行工程设计方法试图在产品的早期设计时期就考虑产品生命周期中的各方面因素,但是要真正实现却是很困难的,该文给出了一个基于多代理的并行工程设计系统框架,来解决“考虑制造的设计（ＤＦＭ）”这一总理２,同样地,并行工程的其它考虑,如可装配性,可维护性与可服务性都能以同样的方式集成到这个框架中来。     

Productivity and the environment

This article discusses the problem of how management can improve the productivity by exerting a closer control over the working environment, so that people suffer less from stresses, for example, as noise, loss of sleep, heat or cold. Laboratory tests alone cannot hope to predict quantitatively the impact of such stresses on the shop floor and in the office; they may, however, be able to indicate in more general terms the kind of shop floor or office which would most benefit from a field study.     

Human‐computer interaction and lean production: The shop floor example

The approach in industrial engineering to computer‐integrated manufacturing (CIM) proved not to be the concept needed for the majority of German companies. At the moment, lean production is the focal point of discussion about production engineering. It tries to combine concepts such as human‐centered organization, internationalization of business, decentralization of decisions, increased development times, total quality management, and process‐oriented business management. There are a variety of implementation areas for lean production. One implementation has been the shop floor‐oriented production support concept as developed by the Fraunhofer Institute IAO in Stuttgart. It focuses on functions such as Computer Numeric Control (CNC) machine tool programming, shop floor control, quality assurance, and resource management, and it aims at decentralized work organization, comprehensive work contents, and support for highly skilled workers on the shop floor. This support is provided by Information Technology (IT) systems that follow the ideas of work enrichment and work enlargement, and use a so‐called shop floor metaphor and graphical user interfaces with standard tool kits. In this way, the shop floor‐oriented production support concept represents a future concept that takes into consideration particularly human factors and human‐computer interaction.     

Customer order management in service oriented holonic manufacturing

One of the most important problems when considering the design of manufacturing systems based on SOA paradigms is the integration of shop floor devices in the business processes at the enterprise level. This paper presents the design and implementation of the Customer Order Management (COM) module based on SOA architecture in the context of holonic manufacturing systems. The COM module is integrating with SOA enabled shop floor devices using industry standards. The implementation leverages a multi agent system suited for industrial applications integrated in a SOA environment capable of dynamic BPEL workflow generation and execution. The prototype consists in a SCA application for core COM module functionality and an extension for NetLogo MAS platform for SOA integration. The COM module interacts with the MES layer using real time events handled by the BPEL process implementation in the execution stage. A web based portal frontend for the COM module has been developed to allow real time tracking of customer orders, providing data about product batch execution and individual progress of each product on the production line.     

A distributed scheduling and shop floor control method

We suggest a market-like framework for scheduling and shop floor control in computer-controlled manufacturing systems where each resource agent and part agent acts like an independent profit maker. A part-resource negotiation procedure is suggested including price-based bid construction and price revising mechanism. Alternative routings for each production order are considered in scheduling and shop floor control. A simulation-based scheduling method is suggested to estimate the start time and the completion time of each task. We develop a prototype soffivare which utilizes the object-oriented concept.     

A manufacturing XML schema definition and its application to a data management system on the shop floor

Digitization for sharing knowledge on the shop floor in the machinery industry has been given much attention recently. To help engineers use digitization practically and efficiently, this paper proposes a method based on manufacturing case data that has a direct relation to manufacturing operations. The data are represented in XML schema, as it can be easily applied to Web-based systems on the shop floor. The definitions were made for eight manufacturing methods including machining and welding. The derived definitions consist of four divisions of metadata, work-piece, process and evaluation. Three divisions except for the “process” division are common to the manufacturing methods. The average number of elements for a manufacturing method is about 200. The represented schema is also used to convey knowledge such as operation standards and manufacturing troubleshooting on the shop floor. Using the definitions, a data management system is developed. It is a Web-based Q&A system, in which the engineers specify the manufacturing case data mainly by selecting from the candidates. Then, the system fills in the blank portions and/or shows messages to help complete the case data. The proposed method is evaluated through practical scenarios of arc welding and machining.     

Implementation of rule-based information systems for integratedmanufacturing

Focuses on the development of a methodology within a software environment for automating the rule-based implementation of specifications of integrated manufacturing information systems. The specifications are initially formulated in a natural language and subsequently represented in terms of a graphical representation by the system designer. A new graphical representation tool is based on updated Petri nets (UPN) that we have developed as a specialized version of colored Petri nets. The rule-based implementation approach utilizes the similarity of features between UPN and the general rule specification language used for the implementation. The automation of the translation of UPN to the rule specification language is expected to considerably reduce the life-cycle for design and implementation of the system. The application presented deals with the control and management of information flow between the computer-aided design, process planning, manufacturing resource planning and shop floor control databases. This provides an integrated information framework for computer integrated manufacturing systems     

MOBMAS: A methodology for ontology-based multi-agent systems development

Ontologies offer significant benefits to multi-agent systems: interoperability, reusability, support for multi-agent system (MAS) development activities (such as system analysis and agent knowledge modeling) and support for MAS operation (such as agent communication and reasoning). This paper presents an ontology-based methodology, MOBMAS, for the analysis and design of multi-agent systems. MOBMAS is the first methodology that explicitly identifies and implements the various ways in which ontologies can be used in the MAS development process and integrated into the MAS model definitions. In this paper, we present comprehensive documentation and validation of MOBMAS.