東京工業大学 大岡山キャンパス 創造プロジェクト館 1階会議室

Toward a frequency-based maritime container assignment model

土木計画学研究委員会 平成21年度第7回国際セミナーのご案内


「Toward a frequency-based maritime container assignment model」
というタイトルで,頻度ベース公共交通配分の古典的モデル(Spiess and Florian,1989,TR-B)を海上コンテナの配分へと拡張適用したご研究の成果について,ご講演いただきます.

参加ご希望の方は,資料の都合もございますので,シューマッカー [ ]まで,ご一報下さい

講師:Michael G H Bell先生
(Professor of Transport Operations,
Center for Transport Studies, Imperial College London)

題目:Toward a frequency-based maritime container assignment model

司会:Jan-Dirk Schmoecker (東京工業大学土木工学専攻客員准教授)

日時:2010年2月19日(金) 午後5時~午後6時15分頃

場所:東京工業大学 大岡山キャンパス
創造プロジェクト館 1階会議室
(東急大井町線 緑が丘駅より,徒歩5分)

On Fri 19 February, 17:00-18:15 Professor Michael G H Bell from Imperial College London will give a seminar titled

“Toward a frequency-based maritime container assignment model”

at Tokyo Institute of Technology. The seminar will be held at Midorigaoka 5th Building, 1st floor meeting room.
Please find below an abstract.
Access to the seminar: Building 20 in this map:

The seminar is open to anyone. For estimation of numbers please send a short reply to me if you are intending to come.

Best regards, Jan-Dirk Schmoecker

In this seminar the classic frequency-based transit assignment method of Spiess and Florian to containers to demonstrate the promise of the method for a global maritime container assignment model. To simplify the problem we consider only one size of container and assume that containers are inter-changeable. To suit the container context, we include the return of empty containers and a capacity constraint for transhipment ports. An origin-destination matrix of full containers is assigned to routes so as to minimise an objective function measured in time units, which considers both full and empty containers. We assume that shipping lines operate strings with given frequencies but base container dwell time calculations on the assumption that vessels arrive randomly and without coordination. Where a port capacity constraint binds, we assume the imposition of a surcharge measured in time units for loading and unloading a container, which is sufficient to bring demand into line wi!
th capacity. We allow the delay to empty containers to be discounted in, or entirely excluded from, the objective function, to reflect the absence of inventory in such containers. Interesting properties of the model are illustrated both by means of a proposition and proof as well as by a simple numerical example. The paper concludes by considering the next steps toward realising a usable container assignment model.