3 boyutlu konteyner yükleme probleminin metasezgisellerle çözülmesi

Container Loading (CL) is a quite interesting and very difficult problem to solve. Given a set of small rectangular items and a rectangular container with known dimensions, the aim is to load the items into the container in such a way that maximum volume utilization of the container is achieved. The Operations Research (OR) literature classifies this problem as NP-hard. Due to the complex nature of the problem, in the first part of this thesis two swarm intelligence (SI) based solution approaches namely Ant Colony Optimization (ACO) and Bees Algorithm (BA) are offered to solve the CL Problem. The results obtained with these approaches are compared with the available approaches in the literature and the performances of these approaches are discussed. Comparison of the proposed approaches in terms of utilization ratio revealed that BA is the best performing algorithm. In addition to this, a CL decision support system - to determine and visualize the packing pattern of a CL problem - is also designed.In the second part, a multi-objective CL (MOCL) problem inspired from a real industrial problem is introduced. The main goal of the MOCL problem is to pack a group of items into the container without any overlap while maximizing the total weight of the packed items and the utilization rate of the container simultaneously. These two objectives are conflicting since the volume of an item is usually not proportional to its weight. The problem is solved via selected multi objective optimization methods (Goal Programming and Weighted-Sum) and the Simulated Annealing algorithm. The proposed algorithms are tested on real data provided by a distribution company and the positive impact of the obtained solution to the company?s transportation policy is discussed.