Path planning for ships in the presence of obstacles

---

Tezin Türü: Doktora

Tezin Yürütüldüğü Kurum: Marmara Üniversitesi, Mühendislik Fakültesi, Bilgisayar Mühendisliği Bölümü, Türkiye

Tezin Onay Tarihi: 2015

Tezin Dili: İngilizce

Öğrenci: DİNDAR ÖZ

Danışman: Ali Fuat Alkaya

Açık Arşiv Koleksiyonu: AVESİS Açık Erişim Koleksiyonu

Özet:

With the increasing number of trade and travel routes over sea and the growing demandof global commerce for containerships, maritime transportation and the operation of anchorage areas gained significant importance. In the safe and efficient management ofmany processes in seaborne shipping, such as path finding under various environmental conditions, path finding with turn constraints and anchorage planning, the means of computer science are being utilized substantially. In this thesis study, three well-known path finding and planning problems in ship navigation is modelled and solved using graph theory and optimization methods. The problems are namely; obstacle neutralization problem, shortest path problem with turn constraints and finally anchorage optimization problem. For the obstacle neutralization problem we proposed Exact Penalty Search Algorithm, that can find the optimum solution in reasonable time for most cases. As for the second problem, a new discretization model for flexible turn angles is presented. It is called Large Adjacency Grids (LAG), which is a generalization of classical lattice. Over LAG we designed an algorithm in order to solve turn constrained shortest path problem after which we demonstrated our algorithm on a ship navigation example in ice-covered waters. Finally we focused on anchorage planning problem. We presented novel metrics to capture safety performance of anchorage planning. Moreover we proposed a new anchorage planning strategy that takes both utilization and safety into account. Our planning strategy achieved upto 98% improvement in terms of safety over two of the recent algorithms. We also modeled nonuniform depth segments of anchorage areas and adapted one of the existing anchorage planning algorithms to our model by which we observed significant improvement in terms of area utilization.