Masatoshi Hatoko

鉄道事業法改正前後で鉄道事業からの撤退がどの程度変化したかを調査し，需給調整規制廃止の影響を分析した．その結果，大手私鉄や第三セクター鉄道などで鉄道の廃線が進行したことが明らかとなった．

ドイツの2都市におけるトランジットモール化されている街路において，歩行者の軌道横断行動をビデオ調査した．この結果，1)軌道横断は大幅に余裕をとる，2)LRTまでの時隔が小さいと横断件数が減少，3)時隔が4秒程度未満で横断しない，4)通過直後に横断，5)通過後も10秒程度は様子見の傾向，などを明らかにした．

LRT 事業成立要件の仮説を 5 つ掲げ、事業が成立している事例と成立していない事例が存在する スペインでの事例を調査することで、その有効性について 検証し、わが国において要件を成立させるための事業上の 工夫などを把握した。

&nbsp;The basic policy for Japanese trunk railway network was established in almost 40 years ago. The second-phase Shinkansen projects based on this policy are nearing completion, but there is no clear vision for future trunk railway network policy including conventional traunk railways. In recent months, MAGLEV project has been just started by Central Japan Railway Company, however, must change the structure of Japanese trunk rail network, but few discussions are held about this matter.<br>&nbsp;In this study, we made a numerical analysis with Genetic Algorithm for searching the best combination of railway constructions and reconstructions at a viewpoint of minimizing of traveling time on trunk railway network under a certain cost constrait. As a result, we found a policy with many small projects of improving trunk lines to be a better policy than with only Shinkansen constructions.

Shinkansen is the only method for developing the Japanese trunk railways under the Nationwide Shinkansen Lines Development Law. However, Swiss Rail 2000 have various methods: implovement of conventional lines, new high-speed lines, and high-performanced cars. The advanced policy achieved success in getting new passengers in spite of their short high-speed lines.<BR>We used GA-based calculation and made it clear that if a Rail 2000 like policy is introduce to Kyushu, how it can perform under a limited construction cost. We concluded that it is more effective to adopt not only constructing high-speed lines but also various methods.

Swiss Rail 2000 policy is a significant policy which has adjusted block times between major stations and timings of train departures on their rail network. In this study, we developed a computer system for searching an effective solution of improvement of intercity railway network considering such policy. This calculation system consists a Genetic Algorithm, because it has to solve a &ldquo;Knapsack Problem&rdquo;: how much and which sectons sould be improved. We worked out some techniques for the problem with a lot of local solutions and for high-speed calculaton of Expected Value of Traveling Time. As a result of sample calculation of trunk rail network in Kyushu, we found that a policy like Rail 2000 may be a new basic rail policy, because smaller investment can bring out bigger effect considering block times and timings of train departures on the rail network.

We paid attention to the relationship between transportation demand for public transport and its frequency, and defined Expected Service Speed as performance index. When demand grows it gets closer to Scheduled Speed. Using unit conversion from car to person, we made the similar relationship of car traffic to be compared with the public transportation.<BR>After that, we examined the interaction between these transportations in urban transit space and found possibilities as followings: Small improvement of public transport cannot cancel traffic jam and cannot increase its passengers in number. Additional lane or decrease in total demand causes outbreak of traffic jam.

This study proposes a new urban-travel demand forecasting system based on a micro-simulation approach as an alternative method to four-step estimation methods. This system comprises, 1) PCATS which simulates an individual's daily activity-travel pattern, 2) DEBNetS which simulates vehicle trips on urban road network, 3) a household simulator which generates future household attributes, and 4) a system to evaluate inter-zonal travel times and costs by public transit based on public transport networks and time tables. The system was applied to replicate the travel demand in a 1990 travel survey data set. A comparison between the observed and the forecast indicated that the system is capable of producing reasonable forecasts.

This calculation system is designed for computing the indexes of traveling time with inter city public transportation such as "Possible Staying Time", "Expected Traveling Time", and so on in the field of study and research. With a correct algorithm, it processes a great deal of public transportation timetable data and works out results for various initial conditions in short time.<BR>If we put data of timetables into this system, it analyze them to a network included relations of stations, airports and train routes. After we select a index and point a start node and its time, we can get numerical values of the index automatically.<BR>This calculation system has already served some studies as data making tool in practice.

In researches which analyze international transportation activities, &ldquo;shortest-traveling-time&rdquo; is often used as an accessibility index. However, it does not appropriately reflect the characteristics of inter-city and international transportation. In this study, we aim at showing the validity of &ldquo;piled-up-traveling-time&rdquo; as an index for analysis of international transportation. We first examine its validity by developing airport choice models and trip generation models. Then we analyze the international accessibility of Japanese cities. From these analyses, we show the change of accessibility caused by transportation project such as Kansai International Airport.