Ant Colony Algorithm For Multi-objective Ship Scheduling Problem

Ship scheduling is an important part of the port scheduling. Ship scheduling optimization problems, which directly affect the efficiency of passenger transportation, service quality and transportation costs, thus affecting the profitability of shipping companies, need to be taken into account. The issues relate to a variety of constraints, such as the demand for passenger seats, the passenger seat of a ship, flight times, sailing speed, crew configuration, and route mileage and so on. At present, the vast majority of domestic shipping (container transport, passenger transport) scheduling still rely on the dispatchers' experience, so that shipping companies can not make full use of transport resources, thus the operating costs are too high. Therefore, in practical work, to do researches on ship scheduling optimization problems, to establish systems that would address these problems so as to improve service quality and resource utilization, are very necessary.This study, originated from the real ship scheduling problem of Shenzhen Xunlong Co., aims to explore an ant colony algorithm for multi-objective ship scheduling optimization problems. It focuses on: how to describe the ship scheduling optimization problem and to establish the corresponding mathematical model under the circumstance of different ship size; how to design the corresponding ant colony algorithm based on the built mathematical model to solve the problem; and how to deal with multi-objective in the ant colony algorithm.The ant colony algorithm designed in the research follows the mechanisms of classical ant colony algorithm in ants' path selection, pheromone update and volatile, combines the method of ant colony algorithm for general nonlinear integer programming in constructing feasible solution space, and uses linear weighting method to transform the multi-objective problem into single objective one, then uses Java language for the implementation of the ant colony algorithm. The experimental results show that the ant colony algorithm can significantly improve the feasible scheduling, with satisfactory results. Finally, this study proposes an ant colony algorithm based ship scheduling decision support system framework, which is of great significance for the present research results' practical application.