| 查看: 150 | 回复: 0 | |||
| 当前主题已经存档。 | |||
kevin.tanqy木虫 (正式写手)
|
[交流]
制造系统控制设计 英文教材 Manufacturing Systems Control Design
|
||
|
Foreword In the late 1980s, the strong needs for modeling, analysis, control, and simulation of complex systems especially computer-integrated manufacturing systems demanded the academic researchers and industrial engineers to seek and investigate better methodologies and tools. Such tools must be able to deal with such system characteristics as asynchronous events, sequences, concurrency, synchronization, mutual exclusion, deadlocks, and choices. While state machines or automata were popular in many applications, they were soon proved to be inadequate since the state explosion problems would be met at the very beginning of system design. Any design flaws or incompleteness may invalidate the entire system design and frequently require rather cumbersome recovery. On the other hand, Petri nets, invented by C. A. Petri in his 1962’s doctoral dissertation, are well equipped with the required capabilities to handle the above-mentioned characteristics. They thus gained their popularity among the researchers of discrete event systems and industrial applications in manufacturing automation. The research group at Rensselaer Polytechnic Institute (RPI) was established and led by Professors Frank DiCesare and Alan Desrochers. It was supported by many leading industrial companies such as IBM, GM, Johnson and Johnson, Sun Microsystems, and Digital Equipment Corporation via an eight-year long Computer Integrated Manufacturing Research Program of the Center for Manufacturing Productivity and Technology Transfer at RPI. They obtained many significant research and application results. Notably, as the first Ph.D. graduate of this group in this area, Dr. Robert Al-Jaar proposed to use generalized stochastic Petri nets for modeling and analysis of production lines. Their work led to their 1994 book Applications of Petri Nets in Manufacturing Systems: Modeling, Control, and Performance Analysis by IEEE Press. As the second Ph.D. graduate of the group, I developed the concepts of parallel and sequential mutual exclusion structures, top-down, bottom-up and hybrid synthesis methods, Petri net-based discrete event controller design and implementation procedures for flexible manufacturing systems (FMS). The results were summarized into the first monograph of its kind, Petri Net Synthesis for Discrete Event Control of Manufacturing Systems, co-authored with Frank DiCesare, Kluwer Academic Publisher in 1993. Dr. Fei-Yue Wang, presently Professor of the University of Arizona and the Institute of Automation, Chinese Academy of Sciences, pioneered in applying Petri nets to designing intelligent machines and building intelligent control foundation together with his advisor, Dr. George Saridis. He also developed a Petri net method for communication protocol design and performance analysis for manufacturing message specification. From the same group, Dr. Inseon Koh, presently Professor of Hong-Ik University, Korea, perfected a bottomup method to synthesize Petri nets with desired properties. Dr. Jagdish S. Joshi conducted performance analysis of network and database transactions in a CIM system. Dr. MuDer Jeng, presently Professor of National Taiwan Ocean University, invented a new class of Petri nets suitable for modeling automated manufacturing systems. Dr. Doo Yong Lee, presently Professor of Korea Advanced Institute of Technology, pioneered in using various heuristics to guide optimal or sub-optimal schedule search in timed Petri net models of flexible manufacturing systems. Dr. Alessandro Giua, presently a professor of University of Cagliari, Italy, developed a supervisory control theory in the framework of Petri nets. Dr. Tiehua Cao and Professor Arthur C. Sanderson combined fuzzy logic theory and Petri nets and developed fuzzy Petri nets for intelligent task planning in a robotic system. The research led to the publication of Intelligent Task Planning Using Fuzzy Petri Nets in the Series in Intelligent Control and Intelligent Automation of World Scientific Publisher in 1996. Dr. Hauke Jungnitz developed approximation methods for stochastic timed Petri nets. Dr. James F. Watson formulated a method for performance analysis of discrete event systems with nonexponential random time distributions and state space estimation of a given Petri net model. The above-mentioned work addressed various issues from model synthesis, performance analysis, simulation, deadlock avoidance, and supervisory control design and made significant contributions to the field of Petri nets and their applications to manufacturing automation. Yet one significant problem remains unsolved: given manufacturing system specifications expressed in Bill of Materials, Assembly Tree, Task Sequencing matrix, and Resource Requirement Matrix, how can one automatically generate a Petri net model and related design for analysis, control, and simulation of a flexible manufacturing system (FMS). This book written by a group of outstanding researchers under the leadership of Dr. Frank Lewis indeed presents an elegant solution to the above long-lasting problem. Their proposed matrix-based approach represents one of the most significant innovations to the area of Petri nets and related discrete-event modeling approaches for manufacturing system control design. The authors are able to identify a unique mapping between the Petri net elements and system specifications and reveal the underlying relations for a number of design and analysis tools used in industrial engineering. More importantly, the research group is able to link what they do to the generation of control code required by Programmable Logic Controllers (PLC). PLC have been the industrial horse in almost every sector of automated manufacturing and packaging industry for three decades. This present book contributes to the area of manufacturing automation in a number of ways. First, it comprehensively presents a matrix-based modeling and controller design framework. It uses an intelligent material handling workcell to illustrate clearly various steps in matrix-based controller design. Second, the book addresses how to utilize matrices for analyzing structural properties of manufacturing systems. It reveals the underlying relationship among graph descriptions, max-plus algebra, and the proposed matrix models. Third, the book investigates a very important yet difficult class of manufacturing systems, namely, multiple re-entrant flowlines. It answers how deadlocks can be avoided in such systems. PLC-controlled flexible manufacturing systems are used to illustrate various deadlock avoidance strategies. Fourth, the book presents Petri nets and their complementary character with the matrix models. A computer-aided design tool called Petri.NET is developed and presented, allowing researchers and engineers to model and simulate FMS using either Petri nets or matrix models. Finally, the book presents the basics of virtual factory modeling and simulation and a number of CAD tools used in industry. Its contribution includes a web tool called FlexMan that can be used to design and simulate of FMS based on virtual factory models and matrix-based methodologies. Such examples as palletization workcell, FESTO FMS, robotic brick-handling system, Volvo body-manufacturing line, and assembly station are used to demonstrate these tools. In conclusion, the authors have well presented their innovative manufacturing control design methods based on matrices, Petri nets and other related discreteevent modeling tools. The book clearly advances the state-of-the-art in the area of flexible manufacturing automation and its impact to the area will last long, not only methodologically but also practically. MengChu Zhou, Ph. D. and Professor New Jersey Institute of Technology Newark, NJ http://web.njit.edu/~zhou [ Last edited by kevin.tanqy on 2008-10-30 at 12:50 ] |
回复此楼» 猜你喜欢
为什么中国大学工科教授们水了那么多所谓的顶会顶刊，但还是做不出宇树机器人？
已经有9人回复
-
版面费该交吗
已经有9人回复
-
体制内长辈说体制内绝大部分一辈子在底层，如同你们一样大部分普通教师忙且收入低
已经有13人回复
-
面上可以超过30页吧？
已经有4人回复
-
“人文社科而论，许多学术研究还没有达到民国时期的水平”
已经有5人回复
-
什么是人一生最重要的？
已经有4人回复