İstanbul varışlı ve çıkışlı yurtiçi otobüs yolculuklarının analizi

Abstract

Bu tezde İstanbul varışlı ve İstanbul çıkışlı yurtiçi otobüs yolculukları analiz edilmektedir. İstanbul varışlı ve İstanbul çıkışlı yolculukların şu anki değerlerine dayanılarak geleceğe yönelik tahminler yapılmaya çalışılmaktadır. Bunun için de şu anki yapılan yolculuklar bir model dahilinde formüle edilmektedir. Bu formülasyon işleminde regresyon analizi kullanılmaktadır. İlk olarak regresyona analizlerine dahil edilecek ve İstanbul varışlı ve İstanbul çıkışlı yolculukları etkileyen parametreler belirlenmektedir. Bu parametreler: yolculuğun başlangıç ve bitiş şehirlerinin nüfûsları ve şehirlerin İstanbul'a olan karayolu uzaklıkları olarak belirlendi. Bunların regresyon analizlerinde direk olarak kullanılması yerine bir model içerisinde kullanılmasının daha uygun olacağı saptandığından bir model oluşturuldu. İki yerleşme merkezi arasında doğabilecek trafiğin hacmini hesap etmek için şimdiye kadar teklif edilen bütün formüllerin şeklinin; P..P." X = K. Dn formunda olduğu görüldü ve bu formülün kullanılmasına karar verildi. Formüldeki Pj yerine İstanbul'un nüfusu, Pj yerine diğer illerin sırasıyla teker teker nüfusları ve D yerine ise illerin İstanbul'a olan karayolu uzaklıkları yazılarak regresyonda bağımsız değişkeni oluşturan X değerleri hesaplandı. Regresyon analizlerinde Y bağımlı değişkeni temsil edecek 1996 yılı İstanbul varışlı ve çıkışlı otobüs sayılarındaki yolcu sayılarının elde edilmesi için ise firmalara anket formları dağıtıldı. Bu formlardaki doluluk oranlarından ve firma yetkinlerinin beyanlarından yararlanılarak normal bir gün için ortalam doluluk oram olarak %70 kabul edildi. Ayrıca, yine aynı şekilde gerçeği daha iyi temsil etmek için yolda inen yol kontenjanı olarak da firma yetkililerinin verdikleri bilgiler sonucunda ortalama %10 seçildi. Toplanan formlardan elde edilen bilgiler bilgisayara girilerek regresyon analizi uygulandı. Ancak, tüm iller regresyona dahil edilerek bulunan regresyon denklemi ve yapılan testler sonucunda yüksek bir korelasyon elde edilemediğinden iller nüfuslarına göre, nüfusu 0-500.000 arasında, 500.000-1.000.000 arasında ve >1. 000.000 olan olmak üzere üç gruba ayrılarak regresyon uygulanması yoluna gidildi. Yapılan her regresyon analizinin sonucunda bulunan değerler teste tabi tutularak regresyon kontrol edildi. Otobüslerin tam dolu olarak gittiği bayram tatilleri, yılbaşı tatilleri, okul açılış günleri ve asker sevkiyatı gibi yolcunun yoğun olduğu zamanlan temsil etmek için tam dolu, normal bir günü temsil etmek içinse %70 dolu kabulü yapıldı. Her bir kabul için ortalama yol kontenjanı olarak seçilen %10 dağıtımlı ve %10 dağıtımsız durumları incelendi. Aynı kabuller hem kış hem de yaz sezonu değerleri için yapıldı. İstanbul varışlı yolculuklarda ise yolcu sayılarının anketlerden elde edilmesi daha zor olduğu için çıkışlı yolculuklara eşit kabul edildi. İki şehir arasındaki illerde inen yol kontenjanı yine ortalama %10 olarak seçildi ve %10 dağıtımlı ve %10 dağıtmısız diye ayrım yapılmadan bütün durumlar için ortalama %10 yol kontenjanı dağıtım uygulandı.

In this study, the inter-city bus travels that leave from Istanbul and come to Istanbul have analyzed. This thesis consists of five chapters. In the first chapter, definition and importance of transportation and importance of thesis are given. The second chapter deals with transportation policies, properties of transportation, factors that effect development of transportation and place of transportation in economy. The third chapter covers the highway transportation in Turkey. The transportation situation in Turkey from the point of view of passenger transportation by bus that forms main subject of thesis has been investigated in chapter four. In the fifth and the last chapter, inter-city bus travels have been examining that connected to Istanbul (leave from Istanbul and come to Istanbul). The main purpose of study is to be estimated bus travels that left from Istanbul and come to Istanbul directed to future. It is made of use bus travels at the present to make this estimation. First, the number of passengers that go from Istanbul to 78 cities of Turkey and go from 78 cities of Turkey to Istanbul was obtained. Investigation forms were distributed to firms in Big Istanbul Coach Station and was wanted to fill for the most density day of week. However, only 240 forms were given back. Essentially, there are approximately among 300-350 firms in 168 offices. This number has been changing day by day. In addition, a big amount of the forms that collected from the firms was observed that were filled carelessly. Therefore, it was given up from being distributed of second form that indicates fluctuating the number of passengers in year. However, various methods were improved to determine the number of passengers in summer months, Christmas, days of school opening, national holidays and soldier consignments etc. These methods were applied due to the impossibility in the collecting the data. Firstly, capacity rate was calculated to each firms from collected forms. Furthermore, it was spoken competent authorities in big firms like Metro, Ulusoy, Varan, Kamil Koç etc. In addition, information of passenger statistics of these firms was obtained in one months. As a result of examining of these, capacity rate for normal a day was accepted 70%. It is impossible to obtain the number of passengers that get off the bus between two cities. Because, this information has not been existed at forms that collected from firms. Therefore, it was spoken competent authorities again. As a result of this, 10% road contingency was determined. Population of cities was taken into consideration to distribute 10% contingency to cities between two cities. XIV For example: POPULATION İZMİR 3.849.188 4772 x 0,10 = 477 passengers get off on the road BURSA 1.603.137 x 2.576.451 477 BALIKESİR : 973.314 y 1.603.137 x_ x = 297, y s 1 80 are obtained. The number of passengers to go to İzmir = 4295 The number of passengers that goes to İzmir fully in one day is 4772. 10% of these passengers have been gotten off Balıkesir and Bursa. According to this assumption, 477,2 passengers are going to balance to 2.576.451 populations that is total population of Bursa and Balıkesir. From here, approximately 10% road contingency was distributed cities between Istanbul and other city. To distribute the number of passengers that come to Istanbul, same method was used. The number of passengers that leave from Istanbul is more logical than come to Istanbul. Therefore, the number of passengers that come to Istanbul was taken to the number of passengers that leave from Istanbul as equal. To determine data for summer months, it was had contact with big firms and other firms in regional area. Amount of journey increases was learned in summer months. Some firms did not want to response. Therefore, the journey increases of these firms were calculated proportionally with other cities that being same region, same size etc. Capacity rate was taken 70% for a normal day and 10% road contingency was distributed cities like winter data. With the assumption of full capacity was represented by the density passenger time like national holidays, soldier consignments, Christmas and day of opening school etc. With the assumption of 70% capacity was represented by a normal day. For summer and winter months : - Leave from Istanbul, full capacity, no 10% distribution, - Leave from Istanbul, full capacity, 10% distribution, - Leave from Istanbul, 70% capacity, no 10% distribution, - Leave from Istanbul, 70% capacity, 10% distribution, - Come to Istanbul, full capacity, 10% distribution between two cities, - Come to Istanbul, 70% capacity, 10% distribution between two cities, Twelve different situations were examined. Some cities that are very close each other, on the same route and the number of passengers are very low were combined. Because, it was foreseen to disconcert regression analysis that will made later. (Table 1) Table 1 Combined Cities To Correct Regression The distribution of 10% road contingency between istanbul and other cities is shown on the table 2. Arrival cities are shown on the first column of the table and cities that is gotten off passengers are shown on the second column of the table. After the number of passengers for summer and winter months and for one day was obtain, these data were formulate to estimate number of passengers in the future. For this reason, regression analysis was used. Firstly, a model was created to include the number of passengers. The parameters that being effected the bus travels was determined. These parameters were selected as : - City population, - The number of workers in the city, - The number of buses in the city, - The number of university students in the city, - The highway distance of city to Istanbul. From these, the number of workers in the city, the number of buses in the city and the number of university students in the city are related to city population. Therefore, these values were not included to regression analysis. As a result of this, regression was examined between city population and the highway distance of city to Istanbul. Table 2 Cities That Were Distributed 10% Road Contingency In Bus Travels That Leave From Istanbul The center of gravity was selected for combined cities. A computer program that is name is "AND" was used for finding highway distance to Istanbul. For example: ANKARA KIRIKKALE (A) (B) Population: 3.236.626 349.396 | 68,9 km- 1 y = 68,9 - x B. 68,9 A + B After cities were combined, the number of cities fall to 63. Population of these 63 cities, the highway distance of Istanbul and the number of passengers was loaded to program of the Microsoft Excel 5.0. It was foreseen that used these parameters in a model instead of directly. This model was determined as R.P,"1 X=K^f- (5.1) In this model, The population of Istanbul was represented by (Pi) parameter. The population of other city was represented by (Pj) parameter. The highway distance to Istanbul was represented by (D) parameter. These X values were used as independent variables in regression analysis. It was observed that the result of regression analysis was not effected with changing K constant. Therefore, K constant was not included in the regression analysis and model. The coefficient of "m" and "n" were determined with test way to maximize the r2 (determination coefficient).