Geri Dön

Demiryollarında boş yük vagonlarının dağıtılması

Başlık çevirisi mevcut değil.

  1. Tez No: 75191
  2. Yazar: LÜTFÜ GÜRAN
  3. Danışmanlar: PROF. DR. ORAL TÜMAY
  4. Tez Türü: Yüksek Lisans
  5. Konular: İnşaat Mühendisliği, Civil Engineering
  6. Anahtar Kelimeler: Belirtilmemiş.
  7. Yıl: 1998
  8. Dil: Türkçe
  9. Üniversite: İstanbul Teknik Üniversitesi
  10. Enstitü: Fen Bilimleri Enstitüsü
  11. Ana Bilim Dalı: İnşaat Mühendisliği Ana Bilim Dalı
  12. Bilim Dalı: Belirtilmemiş.
  13. Sayfa Sayısı: 135

Özet

ÖZET Her türlü ulaştırma sistemi içerisinde yük taşımacılığı hatırı sayılır ölçüde boş yük taşıtı hareketi oluşturmaktadır. Ulaştırma sistemlerinin lojistik planlamasında boş yük taşıtlarının hareketlerinin kontrolü ve planlaması, ulaştırmanın her aşamasında karşımıza önemli bir fenomen olarak çıkar. Bu araştırmada mevcut literatürü inceleyerek; boş yük taşıtlarının hareketlerinin sistematik bir biçimde düzenlenerek, işletme sistemi dahilinde işletme maliyetlerini en aza indirmek amacıyla, problemin bir sınıflandırması yapılmış ve daha önceden dünya literatürü içerisinde boş yük taşıtlarının hareketlerinin düzenlenmesi hakkında yapılan araştırmalar ve bu araştırmalardan elde elde edilen modeller gruplar altında toplanarak, konuya bakış açılan ve içerdikleri algoritmik çözüm yöntemlerinin bir değerlendirmesi yapılmıştır. Ayrıca ülkemiz dahilinde demiryolu yük taşımasını gerçekleştiren Devlet Demiryollarının boş yük taşıtlarının hareketlerini nasıl düzenlediği işlenilmiş ve gelecekte optimum bir fayda sağlanması açısından yararlı olacağına inanılan modeller geniş kapsamlı olarak incelenmiştir. XII

Özet (Çeviri)

SUMMARY EMPTY FREIGHT CAR DISTRIBUTION IN RAILWAYS The transportation of goods is an essential economic activity involving a large number of complex operations for the shipment of multiple products using vehicles of different types and various modes, through networks of a complicated structure. Many transportation systems face the problem of repositioning empty containers or vehicles to points where they can be reload. Air freight carriers, railroads and car rental firms are e few examples of systems in this situation. Many potentials reloading points may exist to which a given empty vehicle or container can be assigned. Unloading and reloading points are often far apart, and the movement of empty vehicles to reloading sites takes time. The empty car distribution problem faced by railroads has received much attention recently as part of the effort to increase freight car utilization. The large number of cars and supply and demand locations involved make it a complicated problem. An average car spends most of its time empty and travels 45 % of its mileage empty. There is reason, then, to believe that car distribution is being performed inefficiently. Railroad cars are expensive and the cost of the distribution process itself is a significant portion of railroad operating costs. A vast literature exists on the many aspects of the problem of empty railcar distribution. Many of these research are mathematically models involving different types of linear programming methods. And many of them are policy models involving stock control systems, freight logistics, network design models, etc. In this study the empty car allocation problem in Railways is taken and it has been tried to understand as a complicated synthesis of some different other components. The study consists of five parts. xmPart One The first part is an introduction and taxonomy of the previous railcar distribution models. This first part includes policy and mathematically models like inventory models, service design network models, the estimation of demands, fleet sizing models, linear programming models and logistics systems design models etc. The policy models are especially long-term models. They could help the car distributors in the long term period. But the mathematically models involving deterministic and stochastic approaches, which are solutions very useful in the short time period. Reweaving with these distribution models it obtained that, two modeling approaches emerge as preferred formulation for the distribution of empty vehicles in the railways. Deterministic approach in the transport models may still be appropriate for situations where, as in the rail mode, travel times between two terminals are of the same magnitude as the length of the planning period. This class of formulations is also very useful for building submodels for more general strategic planning of transport operations. Stochastic models are very useful to estimate the futuristic demands and supplies. To estimate the futuristic demands is very important in the case of railcar allocation, because the most important thing in the railcar allocation is, to know the next step in the case of allocation. If it has been known what the empty railcar demands could be in the near future, it would be prepared yet. The most important and useful model in the field of stochasticity is well explained in the third part of the study. Within the improvements in the linear programming technics in the last twenty years, a new approach is being considered as a mixture of the linear programming models, stochastic models and simulation models. This first and wellknown research of this kind of hybrid models, which is explained in the third part of the study. Part Two The second part of this study includes the most commonly linear programming algorithms Simplex method, Transportation method, The stepping stone method, Cells with minimum costs method, Ford - Fulkerson' s algorithm and the Out -Of - Kilter algorithm. The stepping stone algorithm, transportation algorithm and the cells with minimum cost's algorithm are very useful to get the optimum solution in the case of local and uncomplicated transportation matrixes. These solution technics are very XIVuseful in the static railcar distribution models. The word static represents that the railcar demands and supplies are constant in the distribution period. These linear programming technics could have more difficulty, if the solution is researching in the whole network systems. A large number of the transportation matrixes are formed with the datas from the network and solving these matrixes are a very big problem. The Simplex, Ford - Fulkerson and the Out - Of - Kilter algorithms are very useful algorithms in the computer supported empty railcar allocation models. The most common used algorithms in the linear programming models are the Simplex method and Out - Of - Kilter algorithm. A comparison between these two technics should be made necessary. Simplex method has two stages. In the first stage the possible basic solution should be researched. Regarding this basic solution, the optimum solution will be found in the second stage of the Simplex' s method. But in the Out - Of -Kilter algorithm the basic solution isn't much meaningful, to get the optimum solution. In the Out -Of- Kilter algorithm the two stages, which are discussed in Simplex Case above, are combined each other. The important difference between these two algorithms is the Kilter numbers in the Simplex method are not monotonous. Part Three Section One We could easily see the most important railcar distribution models in the third part of this study. The stochastic - dynamic network model for RailRoad car distribution is explained in the first section of this part. This model assumes a homogenous fleet of railcars. Within the datas from the transportation network, including informations between the origin and destination points and the previous informations about the empty railcar flow, which were occurred in the past; this model could estimate the futuristic demand and the related supply of the empty railcar. The important characteristic of the stochastic model developed here as follows.. The model is dynamic. The railcar distributing decisions in this model considers for the current period supplies and demands in future periods. The planning horizon is at least as long as the longest travel time between the yards in the network.. Uncertainty in forecasts of future and demands is included in the model by the defining the mean and variance in forecasts of the exogenous components of supply and demand at each yard for each period of the planning horizon as model inputs.. As discussed above, travel time variability contributes to uncertainty in future supplies. A component of the model represents the uncertainty in travel time between the yards. XVAnd finally this model expects the profits from : a) revenue from filling orders b) costs from holding cars unused at yards c) stockout costs from not filling orders d) costs of moving cars between the yards This model considers that, the variability of forecasts in future supply and demand and travel time uncertainty is important components of these expected costs and revenues. The model generates daily flows of empty railcars between yards that maximize expected profit. It improves over previous models in two major respects: 1. Uncertainty in forecasts of future supplies, demands and travel times is incorporated into the model structure. 2. A planning horizon is considered in determining optimal daily decisions, with the expected value of empty cars at the end of the horizon included as a function of decisions. All previous railcar distribution models of the network optimization type assume that supplies, demands and travel times are known with certainty. This model the value of cars at the end of the planning horizon a variable, dependent upon car distribution decisions. This is an important step for enabling the model to generate good daily car flow rules. Specific results are as follows: 1. Empty car movements caused by imbalances of supplies and demands decrease as uncertainty increases. 2. Imbalances over the rail network in the expected revenue received for supplying cars to shippers, combined with increasing uncertainty in future supplies and demands, result in more empty cars being moved to, or held at, high revenue yards. 3. Interactions of network characteristics; such as forecast demand and supply levels at each yard, revenue and cost levels, distances and travel times between yards stongly effect the impact of uncertainty. 4. The impact of uncertain future demand is much greater than of uncertain supplies, with travel time uncertainty having only a minor impact on distribution decisions. Section Two The second model is a hybrid model model. It includes stochasticity, linear programming and simulation. This model applies the concept of stochastic linear programming to the homogenous empty freight car distribution problem. This is accomplished by utilizing an algorithm for the distribution of the empty free - running freight cars, which uses linear programming to distribute cars to known freight car Xlllorders and uses stochastic linear programming to preposition empty freight cars to nodes where orders are anticipated. The two optimization problems are linked and driven by the simulation that keeps track of the supply and demand for freight cars in the network. The objective of the two linear programs is to maximize the total transit time of empty cars between nodes. By reducing transit times and the time period between distribution decisions, cars get to the loading points faster. This process increases car utilization and also indirectly reduces the delay of getting empty cars to shippers as measured by late - car - delays. Once demand are known, car - km' s and movement costs are also reduced by the optimization since time and costs are related to transit time between nodes. The results of this model are as follows: 1. A substantial reduction in the final distribution car - hour stands to be gained by prepositioning cars. The greater the amount of prepositioning in the network, the greater the saving of car - hours. 2. Prepositioning works best when the network has a directional demand. 3. Less prepositioning takes place when there are too many cars in the system or when there is a high supply /demand ratio. 4.Prepositioning cars to nodes through which cars would not pass without prepositioning will entail more total car - miles in the system than with the prepositioning. 5. The periodic review period is network dependent and should be judiciously chosen to adequate time for prepositioning. Section Three The third section of the third part includes the empty railcar allocation model, which is prepared for the French Railways. In this model an allocation system, centered on an expert system that reproduces distributors' reasoning, is developed. This system is also enabled the distributors' reasoning process to be understood and modeled, which is a considerable asset for training and improving the performance of each distributor. Moreover, this expert system technology commonly offers an advantage that designers sometimes forget: An editor was able to plan and facilitate updating. Finally, this system improves the consistency of decisions making by different central distributors, which helps regional distributors understand distribution center objectives. But this system has some problems. These problems are as follows: 1. Data Management System cannot provide the exact number of available railcars. 2. Regional distributors often modify their requests at the last minute XIV3. Regional distributors are reluctant to change their practices, which provide them with a comfortable working environment, few responsibilities and little supervision. Part Four In the forth part of the study the empty railcar distributioning in the Turkish Railways is explained. In Turkish Railways, the Center Office of the Movements is the center of the Railcar distributioning process. It controls the empty railcar allocation all over the Turkey. Some offices called, repartition buros, in each management region organize the regional empty railcar allocation under the control of the Center Office of the Movements. The repartision buros obtain information about the railcars movements from all terminals in their region. And within this information they achieve the empty railcar allocation. There is no use of a computing supported system in the field of regional empty railcar distribution in Turkey. Part Five The final part includes the results and conclusions of this study. It is obtained that, the deterministic models could be used in static situations, where the supplies and the demands are known. These models include some solving algorithms like Simplex, Ford - Fulkerson and The Stepping Stone algorithms. But the supplies and demands were not known in most situations. In these situations the distributors should have to predict the supplies and demands in the future. The distribution models that include a prediction in the field of futuristic supplies and demands, are known as Stochastic Models. These models could be effected by dynamic influences like delays in travel times, accidents, network conditions, transport seasons, economical activities, customers' demands and etc. And these dynamic influences constitute uncertainties in the predictions. To solve these uncertainties in the predictions, the previously demands and supplies should be well studied. After the prediction the distribution problem changes from dynamic position to the static position. And the deterministic models could be used easily. The perfect distribution model should include deterministic and stochastic algorithms together. This perfect model should have two main parts. First part is the prediction period and the second is the distribution period where the deterministic algorithms could be used easily. XV

Benzer Tezler

  1. Tez No

    30649

    Hadi Bara'nın sanatsal kişiliği ve yapıtları

    Başlık çevirisi yok

    MEHMET ÜSTÜNİPEK

    Yüksek Lisans

    Türkçe

    Türkçe

    1994

    Güzel SanatlarMimar Sinan Güzel Sanatlar Üniversitesi

    Arkeoloji ve Sanat Tarihi Ana Bilim Dalı

    PROF. DR. SEMRA GERMANER

  2. Tez No

    511592

    Urban regeneration projects from the perspective of social sustainability and public participation: Belgium and Turkey

    Sosyal sürdürülebilirlik ve halk katılımı persfektifi ile kentsel yenileme projeleri: Belçika ve Türkiye

    İZEL BEŞİKCİ

    Yüksek Lisans

    İngilizce

    İngilizce

    2018

    Şehircilik ve Bölge Planlamaİstanbul Teknik Üniversitesi

    Kentsel Tasarım Ana Bilim Dalı

    DR. ÖĞR. ÜYESİ EBRU ERBAŞ GÜRLER

  3. Tez No

    349870

    Urban transformation in meeting places: The cases of Bursa and Yazd

    Toplanma mekanlarında kentsel dönüşüm; Bursa ve Yazd örnekleri

    SIMA VAEZ ESLAMI

    Yüksek Lisans

    İngilizce

    İngilizce

    2014

    Şehircilik ve Bölge Planlamaİstanbul Teknik Üniversitesi

    Şehir Planlama Ana Bilim Dalı

    PROF. DR. AYŞE SEMA KUBAT

  4. Tez No

    106687

    Ortadoğu'da demiryolu ulaşımı

    Railway transportation in Middle East

    AHMET GÜNDÜZ

    Yüksek Lisans

    Türkçe

    Türkçe

    2001

    UlaşımMarmara Üniversitesi

    Coğrafya ve Demografi Ana Bilim Dalı

    DOÇ.DR. NURTEN GÜNAL

  5. Tez No

    356139

    Demiryollarında ray birleştirme yöntemlerinin incelenmesi, alüminotermit ve yakma alın kaynak yöntemlerinin karşılaştırılması

    Investigation of rail joining methods, comparison of aluminothermic and flash-butt welding methods in railways

    FATİH ÖZKUL

    Yüksek Lisans

    Türkçe

    Türkçe

    2014

    İnşaat Mühendisliğiİstanbul Teknik Üniversitesi

    İnşaat Mühendisliği Ana Bilim Dalı

    PROF. DR. ZÜBEYDE ÖZTÜRK

Geri Dön