Stan (Stratejik ulaştırma planlaması modeli) nin Türkiye Demiryolu ağına uygulanışı
Başlık çevirisi mevcut değil.
- Tez No: 55986
- Danışmanlar: PROF.DR. GÜNGÖR EVREN
- Tez Türü: Yüksek Lisans
- Konular: İnşaat Mühendisliği, Civil Engineering
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 1996
- Dil: Türkçe
- Üniversite: İstanbul Teknik Üniversitesi
- Enstitü: Fen Bilimleri Enstitüsü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 83
Özet
The network optimization model that is used to simulate network flows in STAN is a nonlinear multimode multiproduct assignment formulation that minimizes the total generalized system cost. The generalized system cost is applied to links and transfers which are identified in the network and has components of train running costs (monetary units per train-km) and train delay costs (monetary units per train-hour). In order to find train delay costs, a volume/delay function that was given in STAN had been used. Delay functions are most important, since they describe analytically the effect of the traffic intensity on the facility by the estimates of the time necessary to pass trough the facility. Consequently, in models for tactic and strategic planning, capacities are generally not considered as explicit and hard constraints but are instead incorporated into volume/delay functions. Delay functions are defined in yards, at transfers and over the links of the network and thus permit an estimation of the total trip time, from the origin of the traffic to its destination given the capacity and the levels of traffic present on the network. In this study volume/delay functions are applied only to links. For each link a volume/delay function is constructed and these volume/delay functions are calibrated for each situation that is considered. The volume/delay functions are not applied to yards because the avarage service time and capacity may be quite difficult to evaluate precisely in most cases. For yards there is no standart way of optaining a reasonably accurate estimate of capacities. Often only historically observed yard put-through values are available. Consequently, when the yard is not a bottleneck in a given network, nor is it expected to become congested in a future scenario, a constant ( estimated from observed values ) delay function may be used for yards. But unfortunately we, in Turkey don't have such observed values so I was not able to put yard delay functions in my study. Another important aspect of any freight transportation system is the presence of a significant traffic of empty vehicles. The impact of empty vehicle movements on both the operations and planning of a transportation system is very important for the rail mode, since rail vehicles are captive to the infastructure and, thus, the empty car movements take place on the same rail lines with the loaded ones. Consequently the train flows and over the link delays are significantly increased and therefore the proper modeling of stategic rail planning must consider the flows of empty cars. The general procedure that is used to model empty car movements in the network is: a) Define the empty car product and its transportation characteristics b) Estimate the origin destination demand matrix for this product c) Include this demand in the general model and treat it as any other product. XIII
Özet (Çeviri)
No connector links are used, for these links represent the access to different modes of the netvvork. Eight groups of products are assumed to create flovvs in the network. These are solid fuels, ores, metallurgy products, fertilizers, construction metarials, intemational freight and other transportation groups. Each group has its own sub groups. For each product group a typical car was defıned and that each group is exclusively transported by that typical car. This car was defıned by taking the vveighted averege (for each of their main charecteristics) of the car types that most usually transport the product group. For the car types that are going to be used in building the typical car, the car types that are used in Turkish Railways are scanned. The weighted avarage of these car types are then taken in order to fınd the typical car for each product group. The unit costs that are considered in aggregade model of rai l transportation systems, gives for yards, links and transfers, the train and car operating costs in monetary units per train-km and car-km, respectively, the train and car delay costs in monetary units per train-hour and car-hour. in this context, it is important to note that ali unit costs apply to vehicles and convoys. it is neccessary, therefore, to convert our flovvs of product groups (in tons) into corresponding numbers of cars and trains. The weight of a train of product group p, is than taken from the statistics yearbook of T.C.D.D. to convert product flovvs into trains using the formulation given in STAN. The train-km costs, for each group, are computed by using the vveight of trains and a monetary cost for each ton of product they carry. Similarly the train-hour costs are computed using the resulting train-km costs and the avarage train speed. İn the model some hypotheses on the nature of the train traffıc were verifıed for most rail line segments in the netvvork: a)Ali freight trains have equal priority b)Ali trains moving in the same direction, on the same link, travel at the same (and constant) speed c)Consequently, no more than two trains meet at the same time d)On each link, the traffic is globally balanced, meaning that the number of trains traveling in each direction is approximately the same No data on the origins and destinations of the product groups were found when the statistics of T.C.D.D. scanned. Therefore the origin destination matrices of product groups were taken from another study, vvhich was made by I.T.Ü in 1987, by making the necessary modifications in order to agregate the product groups used in that study. xiiThe network optimization model that is used to simulate network flows in STAN is a nonlinear multimode multiproduct assignment formulation that minimizes the total generalized system cost. The generalized system cost is applied to links and transfers which are identified in the network and has components of train running costs (monetary units per train-km) and train delay costs (monetary units per train-hour). In order to find train delay costs, a volume/delay function that was given in STAN had been used. Delay functions are most important, since they describe analytically the effect of the traffic intensity on the facility by the estimates of the time necessary to pass trough the facility. Consequently, in models for tactic and strategic planning, capacities are generally not considered as explicit and hard constraints but are instead incorporated into volume/delay functions. Delay functions are defined in yards, at transfers and over the links of the network and thus permit an estimation of the total trip time, from the origin of the traffic to its destination given the capacity and the levels of traffic present on the network. In this study volume/delay functions are applied only to links. For each link a volume/delay function is constructed and these volume/delay functions are calibrated for each situation that is considered. The volume/delay functions are not applied to yards because the avarage service time and capacity may be quite difficult to evaluate precisely in most cases. For yards there is no standart way of optaining a reasonably accurate estimate of capacities. Often only historically observed yard put-through values are available. Consequently, when the yard is not a bottleneck in a given network, nor is it expected to become congested in a future scenario, a constant ( estimated from observed values ) delay function may be used for yards. But unfortunately we, in Turkey don't have such observed values so I was not able to put yard delay functions in my study. Another important aspect of any freight transportation system is the presence of a significant traffic of empty vehicles. The impact of empty vehicle movements on both the operations and planning of a transportation system is very important for the rail mode, since rail vehicles are captive to the infastructure and, thus, the empty car movements take place on the same rail lines with the loaded ones. Consequently the train flows and over the link delays are significantly increased and therefore the proper modeling of stategic rail planning must consider the flows of empty cars. The general procedure that is used to model empty car movements in the network is: a) Define the empty car product and its transportation characteristics b) Estimate the origin destination demand matrix for this product c) Include this demand in the general model and treat it as any other product. XIIINo connector links are used, for these links represent the access to different modes of the netvvork. Eight groups of products are assumed to create flovvs in the network. These are solid fuels, ores, metallurgy products, fertilizers, construction metarials, intemational freight and other transportation groups. Each group has its own sub groups. For each product group a typical car was defıned and that each group is exclusively transported by that typical car. This car was defıned by taking the vveighted averege (for each of their main charecteristics) of the car types that most usually transport the product group. For the car types that are going to be used in building the typical car, the car types that are used in Turkish Railways are scanned. The weighted avarage of these car types are then taken in order to fınd the typical car for each product group. The unit costs that are considered in aggregade model of rai l transportation systems, gives for yards, links and transfers, the train and car operating costs in monetary units per train-km and car-km, respectively, the train and car delay costs in monetary units per train-hour and car-hour. in this context, it is important to note that ali unit costs apply to vehicles and convoys. it is neccessary, therefore, to convert our flovvs of product groups (in tons) into corresponding numbers of cars and trains. The weight of a train of product group p, is than taken from the statistics yearbook of T.C.D.D. to convert product flovvs into trains using the formulation given in STAN. The train-km costs, for each group, are computed by using the vveight of trains and a monetary cost for each ton of product they carry. Similarly the train-hour costs are computed using the resulting train-km costs and the avarage train speed. İn the model some hypotheses on the nature of the train traffıc were verifıed for most rail line segments in the netvvork: a)Ali freight trains have equal priority b)Ali trains moving in the same direction, on the same link, travel at the same (and constant) speed c)Consequently, no more than two trains meet at the same time d)On each link, the traffic is globally balanced, meaning that the number of trains traveling in each direction is approximately the same No data on the origins and destinations of the product groups were found when the statistics of T.C.D.D. scanned. Therefore the origin destination matrices of product groups were taken from another study, vvhich was made by I.T.Ü in 1987, by making the necessary modifications in order to agregate the product groups used in that study. xiiThe network optimization model that is used to simulate network flows in STAN is a nonlinear multimode multiproduct assignment formulation that minimizes the total generalized system cost. The generalized system cost is applied to links and transfers which are identified in the network and has components of train running costs (monetary units per train-km) and train delay costs (monetary units per train-hour). In order to find train delay costs, a volume/delay function that was given in STAN had been used. Delay functions are most important, since they describe analytically the effect of the traffic intensity on the facility by the estimates of the time necessary to pass trough the facility. Consequently, in models for tactic and strategic planning, capacities are generally not considered as explicit and hard constraints but are instead incorporated into volume/delay functions. Delay functions are defined in yards, at transfers and over the links of the network and thus permit an estimation of the total trip time, from the origin of the traffic to its destination given the capacity and the levels of traffic present on the network. In this study volume/delay functions are applied only to links. For each link a volume/delay function is constructed and these volume/delay functions are calibrated for each situation that is considered. The volume/delay functions are not applied to yards because the avarage service time and capacity may be quite difficult to evaluate precisely in most cases. For yards there is no standart way of optaining a reasonably accurate estimate of capacities. Often only historically observed yard put-through values are available. Consequently, when the yard is not a bottleneck in a given network, nor is it expected to become congested in a future scenario, a constant ( estimated from observed values ) delay function may be used for yards. But unfortunately we, in Turkey don't have such observed values so I was not able to put yard delay functions in my study. Another important aspect of any freight transportation system is the presence of a significant traffic of empty vehicles. The impact of empty vehicle movements on both the operations and planning of a transportation system is very important for the rail mode, since rail vehicles are captive to the infastructure and, thus, the empty car movements take place on the same rail lines with the loaded ones. Consequently the train flows and over the link delays are significantly increased and therefore the proper modeling of stategic rail planning must consider the flows of empty cars. The general procedure that is used to model empty car movements in the network is: a) Define the empty car product and its transportation characteristics b) Estimate the origin destination demand matrix for this product c) Include this demand in the general model and treat it as any other product. XIIIİn order to compare the affect of delays on the network, two scenarios are constructed. The first scenario assumes that no passenger traffic is present on the network and that the service for freight flows takes place in 24 hours. The second scenario assumes that there is passenger trafic on the network and reduces the capacities of the links by assuming the service for freight flows to take place in 12 hours. It could be interesting to analyze the freight flows on links in a future year, say 2000. But there were some reasons that prevented us to make such a study. First, even though the present situation was modelled, there were not enough data to run the model properly (like the estimation of O/D matrices). Second, there was no computer program about the model. All sub-models (delay estimation, empty flow supply and demands e.tc.) and assignments were made step by step using the Excel program, and thus, computations for each scenario had taken much time. After the computations are made for the first scenario 32,224,587,000 TL. of total generalised cost is found. For the second one this generalised cost reaches with the increase of 3,133,160,000 TL. to 35,357,747,000 liras. The change is due to increasing delay on links of the network since, for the first scenario total system delay is 279 hours but for the second one the delay is 598 hours. The increasing delay on links of the network also affects the unit cost per tons. İn the first situation the unit cost chance between the first and eight assignments is 27.8 percent but for the second situation this change in unit post per tons increases to 35.2 percent. XIV
Benzer Tezler
- Pazar odaklılık, rekabet yeteneği ve teknoloji odaklılığın ihracat performansı üzerine etkisi
The effect of market orientation, competitiveness and technology orientation on export performance
FERDA YÜKSEL
Yüksek Lisans
Türkçe
2024
Uluslararası TicaretAnkara Yıldırım Beyazıt ÜniversitesiUluslararası Ticaret ve Finans Ana Bilim Dalı
DOÇ. DR. AHMET SARITAŞ
- Dünyada globalleşme eğilimi ve işletmeye etkileri
Başlık çevirisi yok
SEVGİ KALKAN
Yüksek Lisans
Türkçe
1994
İşletmeİstanbul ÜniversitesiOrganizasyon ve İşletme Politikaları Ana Bilim Dalı
PROF. DR. YILDIRIM ÖNER
- Türkiye'de ihracat yapan KOBİ'lerin ortaklık ve sermaye yapılarının ihracat performansına etkisi
The effect of ownership and capital structure on export performance of Turkish exporting SME's
İLYAS ÇELİK
- Makedonya sorunu: İsim, kimlik ve dış politika
The problem of Macedonia: Name, identity, and foreign policy
REYHAN RAHMAN
Doktora
Türkçe
2022
Uluslararası İlişkilerAnkara ÜniversitesiUluslararası İlişkiler Ana Bilim Dalı
DR. ÖĞR. ÜYESİ ATAY AKDEVELİOĞLU
