Decision Support for Design of Production Planning and Control Systems

Dr. Zora Arsovski CIM center. Faculty of Economic Science, Kragujevac

Dr. Slavko Arsovski

Faculty of Mechanical Engineering, Kragujevac

Summary: In the paper are given the basic elements of CIM company model and the fundamentals of the DSS development by application of the Petri nets. Especially are presented results of investigations that are related to the decision support for design (DSS) of product planning and control systems (PPC). 

Keywords: CIM, DSS, PPC, CIM models, Simulation, Petri nets, Production rules, DSS/PPC

 

The Rider Tool A Real Time Approach to Manufacturing Decision Making

K. Bechrakis, N. Papakostas, N. Giannelos, D. Mourtzis, G. Chryssolouris and A. Fanikos, M. Eleftheriou

Abstract: Distributed decision making is lately considered as a powerful approach for encountering the complexity of the manufacturing environment. In this context, the ESPRIT project RIDER6 has specified a framework for distributed decision making in the production level. This framework is based on a distributed agent architecture that is designed to incorporate the complexity and the detail of a manufacturing system. Two pilot cases are examined in order to evaluate the RIDER concept. RIDER is a three-year ESPRIT IV project that started in January 1996 and is scheduled to end on December 1998.

 

 

A Feature-Based Algorithm for Computer Aided Process Planning of Prismatic Parts

Prof. Dr. - Ing. Habil K.-D. Bouzakis, Dipl. - Ing. G. Andreadis

Laboratory for Machine Tools and Manufacturing Engineering, Mechanical Engineering Department, Aristotels University of Thessaloniki, Greece

Abstract: In the present paper a computer supported procedure to determine the process plan of a work-piece, based on feature recognition and on eligible manufacturing criteria is introduced. The data needed for the recognition of predefined features, are extracted from a neutral file (IGES). The developed procedure provides independence from any CAD program and compatibility and capability for integration into any CAD/CAPP/CAM system, without any restrictions during the design process of the work-piece.

Keywords: Feature recognition, CAD, CAPP, IGES

 

 

 

Estimation of Wheel Life using Neural Network

R. Deivanathan, L. Vijayaraghavan, R. Krishnamurthy

Research Scholar, Associate Professor, Professor Department of Mechanical Engineering Indian Institute of Technology, Madras - 600 036, India.

Abstract: The surface quality problems in the ground component arise mainly due to the deteriorating condition of the working surface of the wheel. Intermittent wheel dressing is the only solution for this and hence a proper estimate of the redressing interval (i.e.) the wheel life, is necessary in grinding process planning. In this paper an analysis have been done to model the wear process, which are largely complicated and hence pose difficulties during in-process implementation. A simple yet effective technique for this problem is the Artificial Neural Network. The performance of neural network in predicting the wheel life is compared with other techniques as the linear regression and the power function model.

 

Empirical Control Strategy for Autonomous Industrial Robot Based on Recognition System and Machine Learning

Zoran Miljkovic

The University of Belgrade Mechanical Engineering Faculty Production Engineering Department 27. marta 80, 11 000 Belgrade, Yugoslavia

Abstract - This paper analysis abilities which modern industrial robot should possess so that it can be called autonomous, owing to development of empirical control strategy. Established concept of empirical robot control at all levels (recognition, sensor-motor coordination and robot control on execution level) is based on learning algorithms of artificial neural networks. Aiming to verify the set up concept of empirical robot control on the basis of the recognition system and machine learning, experimental system is developed. Experimental system consists of SONY TR 512E CCD camera, IBM PC Pentium computer with FAST AVIATOR video card, several developed software subsystems and educational industrial robot MITSUBISHI MOVEMASTER-EX, i.e. anthropomorphic robot Don Kihot. Experimental results showed that empirical robot control strategy, based on empirical control algorithm, successfully realizes manipulation of objects in order to solve robotic assembly task .

Keywords: mechatronics, empirical robot control, recognition system, machine learning.

 

  

Using Formal Description Technique ESTELLE for Manufacturing Systems Specification or Description

Sousa R., Putnik G., Moreira F.

Production and Systems Engineering Department - School of Engineering University ofMinho, Campus ofAzurem, 4800 Guimaraes, Portugal Tel: ++351 (0)53 510278, Fax.: ++351(0)53 510268

Email: rms@eng.uminho.pt Email: putnikgd@eng.uminho.pt ^mail: fmoreira@eng.uminho.pt

Summary: Usually the description of a system is given in natural language or in diagrams. It is very hard to make this kind of informal description clear, concise and unambiguous. Implementations based on informal specifications are usually prone to errors, omissions and incompatibilities. This paper presents an attempt to introduce. Formal Description Techniques (FDT's), originally developed for the area of telecommunications and data transfer, as a mean for the specification/description of manufacturing systems. A brief introduction on standard FDT's, LOTOS, ESTELLE and SDL is given. Several concepts were behind the development of these techniques and this paper exploits the extended finite state machine (EFSM) concept as a suitable form for specification/description of manufacturing systems using FDT ESTELLE.

Keywords: Formal Description Techniques; Formal Specification; ESTELLE; Manufacturing Systems.

 

 

Parallel Kinematic Machines

Sasa Zivanovic,

Production Engineering Department Mechanical Engineering Faculty University of Belgrade / YU

Abstract: - The subject of this paper is the machining system with machine tool, designed by the parallel

mechanism necessary motors, transmission and corresponding control. Such processing system has got, by the rule, six axes and it is kinematic redundant in relation to most methods of cutting which is carried out by traditional machine tools. Incorporated parallel mechanism has got parallel axes, so these machine tools are said to have virtual axes. This virtuality is the consequence of parallel mechanism working principle. This is the first time it appears on such machine tools. Its characteristic is difference between machine axes and necessary tool movements in relation to workpiece in programming. Until present times the unit mapping from necessary movement coordinates in Decart coordinate system on available machine movements was possible, because in traditional machine tools serial orthogonal mechanisms were incorporated. On the basis of insight into already existing accomplishment in realizing machine tools with parallel kinematics and my own researches this paper plan to carry out real knowledge by the method of unfinished concept, which it is better to make self concepts and develop further this conception instead of buying somebody else's unfinished product. That is why designing of certain manufacturing module with parallel mechanism was planned. The idea is to use available resources of traditional manufacturing equipment in an untraditional an educational system without getting into development and building of typical components for communication, programming motors, and control.