Şişeboynu kesimlerde doruk saat akımı için dinamik model yaklaşımı

Başlık çevirisi mevcut değil.

Tez No: 75417
Yazar: SEVGİ ERDOĞAN
Danışmanlar: PROF. DR. ERGUN GEDİZLİOĞLU
Tez Türü: Yüksek Lisans
Konular: İnşaat Mühendisliği, Civil Engineering
Anahtar Kelimeler: Belirtilmemiş.
Yıl: 1998
Dil: Türkçe
Üniversite: İstanbul Teknik Üniversitesi
Enstitü: Fen Bilimleri Enstitüsü
Ana Bilim Dalı: İnşaat Mühendisliği Ana Bilim Dalı
Bilim Dalı: Ulaştırma Mühendisliği Bilim Dalı
Sayfa Sayısı: 104

Özet

Ülkelerin ekonomik ve sosyal gelişimine bağlı olarak artan ulaşım gereksinimi, gerek kentiçi gerekse kırsal kesimlerde, mevcut yol ağlarının giderek bu gereksinimi karşılamakta yetersiz kalmasına neden olur. Hızla artan ulaşım istemi karşısında varolan ulaşım sisteminin yetersiz kalması ise trafik tıkanıklığı problemini doğurur. Trafik tıkanıklığı problemi, özellikle kentiçi ulaşımda kendini göstermektedir. Kentiçi ulaşımda tıkanıklığın en yoğun ve ciddi biçimde yaşandığı kesimler ise yol kapasitesinin önceki kesime göre azalma gösterdiği, geometrik olarak daralan kesimlerdir. Bu kesimler,“şişeboynu”1 olarak adlandırılırlar. Köprüler, tüneller ve şerit sayısının azaldığı yol kesimleri, yol geometrisinden kaynaklanan şişeboynu örnekleri, kazalar ve yol yapım-onarım çalışmaları nedeniyle oluşan daralmalar ise, geçici şişeboynu örnekleri olarak sıralanabilir. Trafik tıkanıklıkları, zaman ve ekonomik kayıplar yanında, çevre ve insan sağlığı üzerinde önemli olumsuz etkilere yol açmaktadır. Bu nedenle, elektronik teknolojisinin de yardımı ile, özellikle doruk saatlerde yaşanan trafik tıkanıklıklarını kabul edilebilir düzeye indirgemek ve yol kapasitelerinin enbüyük etkinlikte kullanımını sağlamak amacıyla türlü çözümler üzerinde çalışılmaktadır. Bu çalışmalar, tıkanıklığın zamana bağlı değişiminin dinamik bir benzetim modeli ile temsili ve tasarlanan çözümlerin model üzerinde test edilerek, karara bağlanması esasına dayanmaktadırlar. Bu çalışmada, doruk saatlerde şişeboynu girişlerinde yaşanan gecikme ve kuyruklanmaları tahmin edecek ve türlü tıkanıklık azaltıcı önlemlerin etkilerinin incelenmesini sağlayacak dinamik bir model incelenmiştir. Bu amaçla, M. Ben- Akiva, A. de Palma ve M. Cyna tarafından geliştirilen dinamik model esas alınarak, bu kesimlerde oluşan kuyruklanmalar ve gecikmelerin en aza indirgenmesi için uygulanabilecek farklı yönetim değişkenlerinin etkileri incelenmiştir. Öncelikle, Dinamik Modelin çıkış noktasını oluşturan Gelişigüzel Denge Modeli incelenmiştir. Modelin işletildiği tek başlangıç-son çiftinden oluşan basit sistem tanımlanmış, bazı temel kavramlar verilmiş ve modeli oluşturan alt modeller kısaca açıklanmıştır. Gelişigüzel Model, Temel Model ve bir trafik kontrol parametresini içeren Geliştirilmiş Model olmak üzere, iki ayrı durumda incelenmiştir. Ardından, Dinamik Model'e geçiş yapılmış ve modelin matematik özellikleri irdelenmiştir. Daha sonra, bu model yardımı ile etkileri incelenebilecek yönetim değişkenleri açıklanmıştır. Son olarak, sabah doruk saatlerinde Boğaziçi Köprüsü girişinde oluşan tıkanıklığın incelenmesi amacıyla modelin bir uygulaması yapılmış ve elde edilen sonuçlar değerlendirilmiştir. Ingilizcedeki“bottleneck”sözcüğünün karşılığı olarak kullanılmıştır.

Özet (Çeviri)

The increasing needs for transportation depending on economical and social development of countries, has caused both urban and rural road networks to disable to hold these transportation needs. For that reason, transportation problems have become more widespread and severe than ever in both industrialized and developing countries alike. General increase in road traffic and transportation demand has resulted in congestion, long delays, accidents and environmental problems well beyond what has been considered acceptable so far. Peak period traffic congestion is a daily occurance in most metropolitan areas, particularly in commuting corridors with bottlenecks. The excessive delay, instability of travel time and increased fuel consumption which accompany congestion translate into significant economic and social cost. Innovative approaches for coping with this problem in urban transportation systems require deeper understanding of the complex interrelation between user decisions - demand side of the system - and the system performance -supply side of the system -. Transportation demand and supply have very strong dynamic elements. Because, flows on transportation networks are the result of complex interaction between user decisions and system performance -demand and supply-. The dynamics of this interaction are of particular importance to study of peak period congestion in cities. Peak period congestion is the one of the most important features of transportation supply. This is a term which is difficult to define. The most simple definition of congestion can be made as inequality between transportation demand and supply. Congestion arises when demand levels approach the capacity of a facility and the time required to use it increase well above the average under low demand conditions. A good deal of the demand for transportation should be concentrated on a few hours of a day, in particular in urban areas where most of the congestion takes place during specific peak periods. This time variable character of transportation demand makes it more difficult -and interesting- to analyse and forecast. Most equilibrium studies focus on a static description of link flows and travel times are invariant over the duration of the peak period. Under those assumptions, users respond to congestion primarily through route choice decisions; however, an important dimension of choice available to users for combating is thus left out, namely the decision regarding departure time. Increasing levels of traffic congestion underscore the importance of traffic management schemes. Although the static approach is successfully used inpractice, it is not suited to address the following questions: What is the impact of a time varying toll? How do flexible and staggered working hours influence traffic congestions? How do driver information systems modify route and departure time choice decisions? There is a need in particular to develop fast simulation tools to evaluate the impact of Advenced Traveller Information Systems (ATIS) which are based on new information technologies which help drivers to make better use of the existing infrastructures. Moreover, congestion cannot be treated adequately using only the total number of drivers using a route during the peak period. Rather an explicit description of the evolution of congestion over time is required. The scope of this study is the problem of peak period traffic congestion and the analysis of alternative congestion relief methods. It presents a dynamic model of the queues and delays at a single point of traffic congestion, desciribed as bottlenecks. Bottlenecks are sections of roadway having a capacity that is less than that of the section of leading up to it. Because of this characteristic of bottlenecks, the major delays to users occur at bottlenecks. The model consists of a deterministic queueing model and a model of arrival rate as a function of travel time and schedule delay. A dynamic simulation model also describes the evolution of queues from day to day. The model is used to study the impacts of changes in capacity, total demand, flexibility of work start time and a traffic control parameter. The model developed in this study considers mainly the problem of the equilibrium between schedule delay and service performance. The schedule delay concept can be defined as follows: To enter the CBD (Central Business District) of many major cities, commuters often have to go through a bottleneck of limited capacity. The two bridges on Bosphorus or tunnels are examples of such a situation. During the peak hours, these facilities are congested, and travel time increases. To avoid congestion and the disadvantage of long travel time a professional worker might choose to arrive at work later, whereas an unskilled worker, who has to be on time, might choose to leave his/her home very early. These two attitudes are examples of schedule delay. The model considers travellers who go through bottlenecks of limited capacity and who may be able to adjust their departure times to avoid slow traffic. Like a commuter who has to be at his/her work place at an official work start time. This desired arrival time may be during the peak period and the road that the commuter take is congested at that time. This commuter may have a choice between on time travel with a long travel time, and a late or an early arrival with a shorter travel time. The difference between actual and desired arrival time is the schedule delay incurred by the traveller who trades it off against travel time. Cosslett, Small and Abkowitz developed econometric demand models of work trip scheduling that are based on this tradeoff between travel time and schedule delay. They estimated logit models of the choice of departure or arrival time with different specifications of the utility functions. The present model employs a probabilistic demand model of the type estimated by Cosslett, Small and Abkowitz. Multinomial logit models are often used in transportation choice models, so, the demand model used in this dynamic model consists of a multinomial logit model. The logit model contains an unobservable random utility ( U(t) ) for an arrival at the entrance to the bottleneck at time t and a deterministic utility component ( V(t) ).The utility function ; U (t) = V(t) + n e(t) e(t) is a disturbance term, assumed to be independently and identically distributed with a Gumbel distribution, n is a scale parameter where ^ = 0 corresponds to the deterministic choice and n = oo corresponds to a pure random choice. Each individual selects the travel time which maximizes his/her utility. Under those assumptions the probability of a given time t being chosen is; t +T 1 P(t) = exp[ V(t) /[i ] / J exp[ V(t) / n ]dt t 1 where the logit denominator is obtained under the assumption that the“ first departure from destination (home) occurs at time 0, which means that the first car arrives at time ti at the entrance of the bottleneck, and that the last possible departure from destination (home) at time T. Thus N cars arive at the entrance of the bottleneck between ti and ti+T. Let E be the logit denominator, t +T 1 E = J exp[ V(t) / n ]dt t 1 and the arrival rate at the entrance of the bottleneck, r(t) = N.p(t) = N/E. exp(V(t)/n The detailed specification of the utility function described in Chapter 2. The dynamic model examined in this study is an extension of the Stochastic Equilibrium Model that was developed by de Palma and Lefevre. This Stochastic Equilibrium Model contains a Basic Model and an extension of it with a traffic control parameter which is named as Extended Model. The dynamic model is an extension of this Sthocastic Model that describes the evolution of the departure time distribution and of the queueing delays from day to day. The Basic Model described in Chapter 2 and analysed for no congestion case in Chapter 3, presents a situation where a queue is forming at the etrance of a bottleneck. A multinomial logit model gives the distribution of departure times. The supply model is deterministic - because, the capacity of the bottleneck per unit of time is assumed to be constant-. By combining the supply and demand models, the Basic Model gives the arrival rate at the entrance of the bottleneck and the evolution of the queue as a function of beginning and ending times of the congestion period. An extension of Basic Model that includes a traffic control parameter, is given in Chapter 4. This extension assumes that arrival flow is spread on a portion of the road before the bottleneck. As before, the speeds on roads before, after, and on the bottleneck are fixed and the travel times are constant. However, the speed on this section (tu) depends on the number of vehicles on it. The only difference between the Basic and the Extended model is this parameter, which is demonsrated by tu. This additional parameter removes the property of thebasic model that congestion occurs as soon as the arrival rate r(t) exceeds the capacity of the bottleneck. In the Extended Model, congestion can be avoided with a sufficiently large t”that can“absorb”the fluctuations of arrival rate r(t). It is possible to consider tu as a parameter of a traffic control device that regulates the arrival rate at the entrance to the bottleneck. Using a system of traffic lights, for example, the arriving traffic is uniformly distributed over this deterministic queueing segment (tu) and thereby decreasing the intensty of congestion. If tu is large enough, no congestion will appear. The importance of traffic control is also to eliminate small fluctuations and produce a smooth arrival rate. In Chapter 5, the main results of the Sthochastic Equilibrium Model are given and the dynamic extension is described. Before describing the dynamic model specifications a general framework for Dynamic Network Equilibrium Models is given. This framework deals with the dynamic nature of network performance under time varying demands. It will consider a related and equally important topic in demand modelling concerned with the dynamic day to day adjustment processes that result in transient disequilibria. In Chapter 6, the Dynamic Model specifications and basic equations are described and also the policy variables and their possible effects are defined. The setting of the Dynamic system is the same as the Stochastic Equilibrium Model, with additional notation to indicate day to day variability. Thus, it also predicts the transient impacts from the implementation of a new policy. Moreover, a simulation program developed and employed to give numerical results. Because, both models -Basic and Extended Models- can be completely solved analytically when the beginning time of the congestion period tq is known. However, tq cannot be known, thus a numerical solution technique is necessary. The dynamic simulation starts from a situation which in this study is assumed to be the no congestion state. The vehicle arrival rate r, number of the vehicles in the queue D are computed and the waiting times are updated which is now the starting point for the following iteration. Each iteration can be looked at as represanting a day. The nth iteration is the system's state n days after the initial state. The process stops when the relative difference between one day and the next is smaller than a given tolerance. In Chapter 7, a numerical example is employed to see the effects of policy variables on peak period traffic congestion. In the example, Anatolian side of the Bosphorus Bridge is taken as origin and European side of it is taken as destination. Simulation examples was both performed with the Basic and the Extended Models for the following main variables; Total number of the cars (N) Capacity of the bottleneck(S) Range of on time arrivals(A) Scale parameter of the choice model(^) Traffic control parameter^) 23000 cars 7200 cars/hr ±0.5 hr 1.2 0 (for the Basic Model) 0.09 hr (for the Extended Model) In this example, the initial conditions of the dynamic simulation was a situation without congestion which is reflected in a flat departure rate distribution during theon time arrival interval. In the Basic Model, congestion period begins when the departure rate exceeds the capacity of the bottleneck, whereas in the Extended Model it begins whenever the number of cars in the queue is greater than S.tu. So, in the Basic Model, the departure rate distribution moved rapidly through to beginning time of the day which is not a consistent situation with reality. According to the simulation results, the Extended Model's results were more realistic than the Basic Model. For that reason, Extended Model was used in the simulation experiments. In the simulation experiments, the values of the policy variables are changed one at a time. For the steady state solutions, the outputs are : beginning and end times of the congestion, on time arrival interval, delays (maximum and average delays), departure rate and number of the cars in the queue. The results were presented with graphs that describe summary stationary state -or the last iteration results when the stationary state cannot be reached- outputs as functions of the policy variables. The most important graphs are for: the average waiting time, the maximum possible delay, the beginning and end times of the congestion period. The policy analysis graphs were drawn for the following ranges of values: N/S = 1.92 to 3.93 A = 0.10 to 0.55 hr tu = 0.02 to 0.14 hr (fortu+to =0.16) The results obtained from the simulations can be summarized as follows. The average delay and the maximum waiting time have similar behaviour : they both decrease when the ratio of number of cars over capacity decreases, when the flexibility of work start time increases and when the level of traffic control increases. Thus attempts to reduce the average delay also lead to lower maximum waiting time. The duration of the congestion has a different behaviour : it stays constant when the flexibility of work start time increases, and decreases when the ratio of number of cars over capacity decreases and when the level of traffic control parameter increases. The data used for this simulation experiments is not detailed and the model itself is developed for a simple system; so, it can be said that, it is not enough to simulate delays and queues at the entrance of the Bosphorus Bridge. Despite these deficiencies, it is believed that such a study will be useful to analyse the effects of different congestion relief methods -policy variables- on congestion. In the last Chapter, the main results obtained from this study and the deficiencies of the models are disscussed. Moreover, some possible extensions are considered. These extensions are namely; peak-period pricing, market segmentation and mode competition. The useful results obtained from this study can be summarized as follows. The following variables determine the congestion pattern : number of cars in the system, system capacity, flexibility of work start time, traffic regulation, peak period pricing. In order to evaluate a policy, a decision maker has to know what the effects on the level of congestion of a change in one or more of these variables are. This study shows that if the dynamic model in this study can be developed for a large network -a more complex and realistic model-, it can be used as a policy evalution tool.

Benzer Tezler

Tez No
381854
Kent içi otoyolların katılım kesimlerinde deterministik ve stokastik yaklaşım ile kapasitenin incelenmesi: İstanbul örneği
The capacity evaluation of the merge sections in urban freeways considering deterministic and stochastic approach: Example of İstanbul
FULYA ÖZSAN
Yüksek Lisans
Türkçe
2015
Ulaşım İstanbul Teknik Üniversitesi
İnşaat Mühendisliği Ana Bilim Dalı
DOÇ. DR. KEMAL SELÇUK ÖĞÜT

Geri Dön