Çok modelli/ürünlü montaj hatların dengelenmesi için yeni bir model ve çözüm yöntemi
Başlık çevirisi mevcut değil.
- Tez No: 75048
- Danışmanlar: DOÇ. DR. MEHMET TANYAŞ
- Tez Türü: Doktora
- Konular: Endüstri ve Endüstri Mühendisliği, Industrial and Industrial Engineering
- Anahtar Kelimeler: Montaj hattı dengeleme, Assembly line balancing
- Yıl: 1998
- Dil: Türkçe
- Üniversite: İstanbul Teknik Üniversitesi
- Enstitü: Fen Bilimleri Enstitüsü
- Ana Bilim Dalı: Endüstri Mühendisliği Ana Bilim Dalı
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: Belirtilmemiş.
Özet
Bu çalışmada, çok modelli/ürünlü montaj hatlarının dengelenmesine yönelik yeni bir yöntem tanıtılmıştır. Bu yöntem geliştirilirken ve denenirken literatürde yer alan iki çalışmadan yararlanılmıştır. Tez çalışması; giriş, beş tane temel bölüm ile“Sonuçlar ve Öneriler”bölümü olmak üzere yedi bölümden oluşmaktadır. Giriş bölümünde, montaj hatları konusundaki gelişmelerin ve yapılan çalışmaların çok genel bir çerçevesi çizilmiştir. İkinci bölümde genel anlamda üretim kavramından sözedilmiş, üretimin tarihsel gelişimi ve farklı durumlara göre üretim sistemleri anlatılmış, akış hatları sınıflandırılmıştır. Üçüncü bölümde montaj hatları ve hat dengeleme konusuna girilmiş, hat dengelemeyi etkileyen temel etmen ve kısıtlardan sözedilmiş, dengelemede kullanılan temel kavramlar örneklerle açıklanmış, montaj hatları ve dengeleme yöntemleri sınıflandırılmıştır. Dördüncü bölümde tek, çok ve karışık modelli/ürünlü montaj hatları konusunda geliştirilen matematiksel ve sezgisel hat dengeleme yöntemlerinden ve yapılan diğer çalışmalardan sözedilmiştir. Ayrıca konum ağırlıklı dengeleme ve 0-1 tamsayılı programlama teknikleri, yeni yöntemimizin geliştirilmesinde ve test edilmesinde yoğun olarak kullanıldıkları için birer örnek problem ile tanıtılmışlardır. Beşinci bölüm, bu çalışmanın özgün kısmı olan, geliştirilmiş yeni yöntemi içerir. Bu yöntemin genel yapısı ve hangi varsayımlar altında tanımlandığı açıklanmıştır. Yöntemin algoritması ve akış diyagramı ile küçük ve büyük boyutlu örnek problemler üzerinde yapılmış uygulamaların sonuçlan verilmiştir. Altıncı bölümde, yeni yöntemin bir konfeksiyon fabrikasında yapılan uygulaması ve elde edilen sonuçlar sunulmuştur. Bu tez çalışması, araştırma sonucu elde edilen bulgular ve bu konuda ileride yapılabilecek çalışmalar için önerilerin bulunduğu Sonuçlar ve Öneriler bölümü ile sona ermektedir.
Özet (Çeviri)
During the industrialisation, it was thought that a job can be divided into the work elements done by several operators, so that production can be made faster and cheaper. As a result, production is made on a line that moves from one workstation to another. An assembly line is a sort of line on which materials are transferred from one station to another, and operator/operators work on materials based on several constraints such as precedence conditions and cycle time. Operators at the workstations placed on the assembly line perform their job/jobs once the materials/semi-products are received by them. As a result, a completed product is obtained at the end of the assembly line. While designing an assembly line with one or more products, balancing the processing times for multiple lines becomes a problem The aim is to leave minimal or no wasting time to each operator during the assembly after assigning work elements to the workstations. In other words, the final objective is to minimize the difference between the times spent at workstations. At this point, line balancing can be a problem. Assigning work elements to workstations with minimal loss is called“assembly line balancing”or“line balancing”. Increased production rate, effective planning and solutions to the financial problem can be achieved in industry by balancing assembly lines. Assembly lines are important sub-systems of mass production. Raw materials or semi-products can enter to the line in the beginning or at interval workstations. A whole product is obtained at the end of the assembly line after being processed at workstations. The objectives of building an assembly line can be given as follows: obtaining a regular material flow, using the operators and machines at the maximum capacity, performing work elements in possible shortest time, minimizing the loss time, minimizing the number of workstations on the line, assigning work elements to workstations with minimal loss, and minimizing the manufacturing cost. But, it may not be possible to achieve one of these objectives. It should be noted that the main objective is to obtain an optimal result with minimal assembly cost. Today, assembly line balancing is a classical subject of production planning and control. It is one of the problems that is faced very often in practice. When designing an assembly line with one or more products, we should aim to make the total time for each workstation close to each other or exactly the same. The finalobjective is to minimize the difference between the times assigned to each workstation. At this point, line balancing can be a problem. Line balancing can be defined as assigning work elements to workstations with minimal balancing loss. Assembly lines can be divided into three main groups based on the number of the products or models: 1) Single model assembly lines. 2) Multiple model assembly lines. 3) Mixed model assembly lines. Single model assembly lines have been studied extensively in the past. Since 1954, several mathematical and heuristic methods have been developed for single model assembly lines. In 1967, studies started focusing on multiple model assembly line problems. However, there are only a few studies for mixed model assembly line balancing. The new method introduced in this study is a step towards filling the information gap on multi-model assembly line balancing. Today, customer satisfaction plays an important role in determining the marketing and manufacturing strategies. This results in an increase in the number of products/models manufactured and a decrease in the response time to customer demands. Because of this, multi-model assembly line balancing is an important and critical subject in the manufacturing and service industries. Another important subject for companies is to reduce the manufacturing costs. Decreasing the cycle time and increasing the number of models on line may result in an increase in the manufacturing costs. The main objective is to reach the intended cycle time at the optimal cost level. The new method introduced in this study has been developed by taking this point into consideration. The number of workstations (workmanship and machine processing time) and balancing loss (operator idle time and machine idle time) are two important factors affecting the manufacturing cost. In this study, a new method for multiple model assembly line balancing has been developed. When developing and testing this model, mainly two studies in the literature have been used. There are seven chapters in this study including introduction, main chapters, results and recommendations. In introduction, relevant literature is reviewed with an emphasis on recent improvements in line balancing. In Chapter 2, it is aimed to define the subjects of production, production management and their historical developments. Also, production systems and flow lines are categorized and the relevant examples are provided in this chapter. In Chapter 3, assembly line, line balancing, the objectives of the line balancing, main factors and the constraints influencing line balancing have been defined. Main concepts used in line balancing have been described with relevant examples. The influence of the number of models on line balancing is also described in this chapter.The simple and general assembly line balancing problem is described. Finally, classification of the models for assembly line balancing is provided. In Chapter 4, the mathematical and heuristic methods that were developed for balancing assembly lines with single, multiple and mixed models have been summarized in chronological order. The relationships between the models are described. Heuristic, exact procedure, branch and bound, 0-1 integer-programming, shortest path, dynamic programming and work sequencing, parallel station and precedence matrix, deterministic, probabilistic and stochastic line balancing methods have been described in this chapter. The methods to determine the optimal cycle time or the number of stations are presented. Also, the methods trying to achieve other objectives (minimizing the manufacturing cost or learning time etc.) and the simulation models are summarized. Also, other relevant studies and simulation models are presented in this chapter. Positional weight and 0- 1 integer programming techniques have been explained by relevant examples since these techniques have been used extensively when developing the new method. The new method is presented and discussed in Chapter 5. First of all, the problem has been described. At this point, it is useful to give the detailed information about the problem In this problem, multiple model line balancing system has been examined. For each model, cycle time and technological precedence diagram are given. Parallel workstations (additional operators) can be used to reach the intended production rate. If the same work element with different times is used in different models, each work element must be numbered separately. The assumptions made when developing the new method are as follows: all work times are deterministic; any work can be done on any machine; transportation times from one machine to another are assumed to be zero (0); there are no stocks in the system; there are no machine failures or work accidents in the system; there are no product defects. After that, the steps of the new method are given. In the new method, first of all, each product/model has been assumed to be an assembly line model to single minimum number of workstations required for balancing is found. Secondly, each product is separately balanced by using the positional weight balancing technique. The maximum number of workstations obtained is selected. Thirdly, balancing is done for the exact (integer) numbers that are between the minimum and maximum numbers obtained at the previous step. The sequence of the work elements determined at the second step is used. Finally the optimal solution is the one that requires the minimum number of operators. If there are more than one solution, the optimal solution is the one that results in minimum balancing loss. The philosophy of the new method and its difference from other methods have been presented in this chapter. The algorithm and the flow diagram for the new method have been provided. A new rule used for determining the precedence when balancing the assembly line the second time and the test results are presented. Then, the new method has been applied on a small problem.It should be noted that a new approach has been taken when developing the new method. This approach is to convert a multiple model to individual models. It is important that the number of individual models must be equal to the total number of models given with the multiple models. After that, re-balancing of each assembly line must be done for the numbers of workstations which are between the smallest theoretical number and the largest workstation number obtained after the first balancing. Twenty sample problems have been created manually by using all possible combinations of four models (2, 3, 4 and 5) and five work element numbers (10, 15, 20, 25 and 30). A computer program (Program I) is used to solve all the problems created. The results are reported in tables. Also, a new computer program (Program II) is developed to create three hundred problems. After that, these problems have been solved by using Program I and results are presented. A 0-1 integer-programming model for a group of twenty problems has been developed. This model has been solved by using Hyper-Lindo-1989. The results have been compared with the results obtained after using the new method. All the results and the method have been analysed at the end of Chapter 5. In Chapter 6, the results of using the new method in a textile company manufacturing five different models of shirts with similar work elements on a single assembly are presented. The final chapter presents conclusions of the research and points out directions for further research. The developed multi-model assembly line balancing methods have used the combined technological precedence diagram. Works elements assigned to each workstation must be the same when using the combined technological precedence diagram. Whenever there are several models with common work elements, these work elements must be taken care of by operators at the same workstation/workstations. However, in practice, work elements can move from one station to another based on the model manufactured. Also, parallel workstation usage became very common in general. In the new method introduced, each model is balanced based on its technological precedence diagram. During the set-up time for moving from one model to another, the common work elements can be transferred to different workstations. This process results in reduced idle-time costs. The new method gave very close results to those of 0-1 integer programming. Indirectly, workstation set-up costs are at the optimal level. When using the 0-1 integer programming method, solution time is increased exponentially with the increasing number of models/products. In comparison, the new method requires less solution time than that of the 0-1 integer programming method. Also, the combined technological precedence diagram must be prepared for setting-up the models. However, in the new method, each model is balanced based on its own technological precedence diagram.\ As a result, multi-model line balancing problem is modeled very close to real-time conditions. The new method gave very close solutions (better balancing loss and less cost) to the optimal solution. In addition, this method gave better results for large- scale problems in terms of modeling and solution time.
Benzer Tezler
- İmalat sanayinde KOBİ'lerin büyük sanayiyle entegrasyonunun incelenmesi ve Japon modelinin uygulanabilirlik araştırması
Investigation of integration between SMIE's and big industry im manufacturing industry and research of Japanese model's adaptation
HAKAN ŞENOL
- A Discrete time quası birth and death model of fixed cycle time policies for stochastic multi item production inventory problem
Çok ürünlü rassal üretim envanter sistemlerinde sabit çevrim zamanlı politikalar için kesikli zamanlı yarım doğum ve ölüm modeli
AYŞE KOCABIYIKOĞLU
Yüksek Lisans
İngilizce
2000
Endüstri ve Endüstri MühendisliğiOrta Doğu Teknik ÜniversitesiEndüstri Mühendisliği Ana Bilim Dalı
DOÇ. DR. REFİK GÜLLÜ
- An Aggregate production planning and machine loading application at a multi-item batch production plant
Çok ürünlü yığın üretim yapan bir işletmede toplaşık üretim planlama ve makina atama uygulaması
JAMSHİSİ ABBAS MIRAN
Yüksek Lisans
İngilizce
1993
Endüstri ve Endüstri MühendisliğiOrta Doğu Teknik ÜniversitesiEndüstri Mühendisliği Ana Bilim Dalı
DR. NUR EVİN BOZŞAHİN
- Çok ürünlü üretim sistemlerinde amaç programlama yaklaşımı
Goal programming approach to multi product production systems
BAHAR ÖZYÖRÜK(BUMİN)
Yüksek Lisans
Türkçe
1992
Endüstri ve Endüstri MühendisliğiGazi ÜniversitesiEndüstri Mühendisliği Ana Bilim Dalı
Y.DOÇ.DR. SERPİL EROL
- Çok ürünlü tesislerde optimum üretim sıralaması
Optimum production scheduling in multi-product plants
ZAFER ÖZDEMİR
Yüksek Lisans
Türkçe
2000
Kimya MühendisliğiAnkara ÜniversitesiKimya Mühendisliği Ana Bilim Dalı
PROF. DR. RIDVAN BERBER