Tez özetleri Astronomi ve Uzay Bilimleri Anabilim Dalı

Fuzzy Logic Modelling Of Fleet Management

Sea transport forms an important mode and an alternative for the existing transport systems for Istanbul which is surrounded by sea by three sides and having had increasing traffic problems. Maritime logistics is one of the most economic mass transportation systems.

Nowadays, it is seen that even the land-locked fly flags at sea. Maritime logistics, which grants means of mass transportation, as well as being economic for both internal and external trades, is very important for Turkey consisting of two big peninsulas. It is extremely impotant to make an efficient timetable in order to use both ships and quays. Making timetable for ships is called ship scheduling in its literature.

The number of passengers in and out of ships, the quay constraints, passenger capacity of the ships and the other specific constraints of conveyance should be considered when constructing a ship schedule. As the cost of constructing a new quay is very high and usually only one ship can be alongside to a quay, it is very important to determine the waiting time at the quays. The existing schedule is constructed by specialists. In this study a new schedule is constructed in which the route is known. In the new schedule, without changing the last quays’ arrival time, the waiting time in the quays are decreased. The investigations showed that the waiting time in the quays should be determined according to the number of passengers transferred and the quay constraints in each quay.

The calculation of waiting time in the quays were done by considering the number of passengers and the quay constraints. Beacause of the uncertainity of the effect of the number of passengers on waiting time in quay, fuzzy logic modelling is used. The number of passengers in and out of the ships are inputs, and the waiting time is the output of the model. Real time measurements were done in order to understand the relation between input and output layer variables. These real values are used to construct the rule base in the model with Wang-Mendel Neuro-Fuzzy Rule Generation Method. The waiting time in the quays were calculated with this model. The obtained values were used in the new schedule in order to use the quays and ships more efficiently.