Hücresel imalat sistemlerinin tasarımı ve kontrolü için benzetim amaçlı bir üreteç

Abstract

imalat sistemlerinde ortaya çıkan geniş ve karmaşık problemlerin modellenmesi ve çözülmesinde benzetim dilleri vazgeçilmez karar destek araçları olmaktadır. Karmaşık sistemlerin benzetim tekniğiyle modellenmesi sayısız avantaj sağlamaktadır. Ancak, bu teknikle gerçek bir sistemin bir modelinin yaratılması kolaylaşsa bile geçerli bir model kurmak verilecek çaba hiç de küçümsenemez. Bu, benzetim dillerinin özel problemler için tasarlanmamasından kaynaklanmaktadır. Bu tezin konusu hücresel imalat sistemlerinin tasarımı ve kontrolü için benzetim amaçlı bir üreteç geliştirmesi ve benzetim deneylerinin üreteç aracılığıyla gerçekleştirilmesidir. Tez boyunca sırasıyla hücresel imalat sistemleri, tasarım aşamaları, problemleri, benzetimi, SIMAN benzetim dili, üreteçler, literatürdeki çalışmalar ve üreteç geliştirme aşamaları anlatılmıştır. üretecin oluşturulması sırasında, hücresel imalat sistemlerinin karakteristiklerini yapısında toplayan genel bir benzetim modeli geliştirilmiş ve bu model hücre sayısı, iş gören sayısı, iş görenlere atanacak tezgah sayısı ve parça sayısının değişken olduğu, farklı operasyon sıralarına ve farklı operasyon sürelerine sahip "herhangi bir" sistemin tasarımına kadar genişletilmiştir. Tabii ki "herhangi bir" sözcüğünün taşıdığı anlam, tüm benzetim çalışmalarında da yapıldığı gibi, varsayımlarla hafifletilmiştir. Son bölümde ise üreteç kullanılarak bir hücresel imalat sisteminin SIMAN benzetim modeli kurulmuş ve çalıştırılarak sistem performansını veren ölçütlerin değişimi incelenmiştir.

A Simulation Generator for Cellular Manufacturing Systems Design and Control Through the last two decades, cellular manufacturing has emerged as an important manufacturing phylosophy. The implementation of cellular manufacturing does not depend on automation. Thus it can be applied even in classical production systems. Benefits from cellular manufacturing realized due to use of setups, reduced material handling, and the use of manufacturing resources. It is based on small batch sizes and large variety of part types. This requires identification of groups of machines which can produce parts with similar processing requirements. Machines in each cell has been placed close to each other to process part of families. Part families assigned to cells satisfy those two objectives: - Minimize the total material handling. - Maximize the machine utilization. Several approaches have been developed to identify part families and their associated machine cells. Howevwer a substainal number of problems still remain: - xii - First, multitude of clustering algorithims, clustering criteria and measures of performans make it difficult to evaluate and select an appropriate or better clustering method. Second, although it is clear that different clustering criteria may produce different grouping results even if the same algorithim and data are used. And question "which one is best?" is still in a contention. Third, although minimizing the number of exeptional element has been widely used as a measure, the question "does any other better measure exist?" remain to be answered. Examining the responses of such problems of cellular systems under different circumstances have been enabled by computer simulation. Recent developments in simulation languages v have enabled the modelling of complex and large systems. System modelling using simulation has many advantages. Even though the creation of realistic models using a simulation language is become easier, it is not trivial. Because simulation languages are not designed to the problem spesific, so that, modelling of systems can be quite time consuming. The objective of this master thesis is to develop and test A Simulation Generator for Cellular Manufacturing Systems Design and Control. A simulation generator is an interactive software tool that translates the logic of a model described in a relatively general symbolism in to the code of an existing simulation language and so enables a computer to mimic behavior. The generator converts data describing the cellular system into simulation model and automatically runs the simulation program. - Xlll - Chapters and their contents can be given as follows: The first chapter includes general descriptions and definitions of some concepts which were used through the thesis. These are cellular manufacturing system and its characteristics, problems of cellular manufacturing systems, application of simulation in manufacturing systems and simulation generators. Design steps of cellular manufacturing systems, clustering methods and man - machine interference problems were included as a structure of the second chapter. The widely used clustering algorithims were clasified and an important problem of manufacturing systems: "Assigning machines to the operators" was described through the chapter. SIMAN simulation language and simulation models of cellular manufacturing systems are included in chapter three. It was developed SIMAN model macro structure identifying elements of the cells in chapter four. That is performed after a study that takes eight months. Chapter five includes an implementation of simulation generator. Model system consist of four cells including 16 machines and 7 workers which are processing 17 part types. Simulation experiments have been executed for ten days (4800 min) and simulation results have been obtained from SIMAN output processor. In the final chapter, it was sugested following ideals reletad to using of simulation generator for further studies and researchs. Comparing different clustering methods. Examining the effectiveness of process planning by monitoring of cell performances. - XIV - Calculating number of machines asigning to one operator. Examining of part sequencing rules. Determining the ideal conditiones for cellular manufacturing systems. Comparing other manufacturing systems and cellular manufacturing systems. Finally, developing an expert system which gives an explanation of simulation results and statistical suggestions by using interference mechanism. It was also included responses of different systems for various production conditions overall the study. On the other hand, great help was supplied in modelling by using of the simulation generator. In this study, SIMAN was used as an existing simulation language. SIMAN modelling framework is based on two structure: Model and experimental frame. The system model defines the static and dynanic characteristics of the system. In comparison, the experimental frame defines experimental conditions under which the model is run to generate specific output data. By seperating the model structure and experimental conditions into two element, different simulation experiments can be run by changing only experimental frame. The system model remains the same. Anhother helpful aspect of SIMAN is capability of using either discrete event or continuous modeling aproachs or both of them. It was followed three general steps to develop the.simulation generator : - XV - First, a general framework for the systems of interest was defined. This steps includes the identification of the major components of the cellular manufacturing systems (number of cells, machines, workers, transportation devices, parts and part routing informations). Secondly the general framework was coded using SIMAN simulation language. The simulation generator was created and coded at the third step using QUICK BASIC 4.5 general programming language.