34. Evolving relocation rules for the Container Relocation Problem using Genetic programming
Invited abstract in session WD-4: Large scale optimization and applications 2, stream Large scale optimization and applications.
Wednesday, 12:00 - 13:30Room: C105
Authors (first author is the speaker)
| 1. | Mateja Đumić
|
| School of Applied Mathematics and Informatics J. J. Strossmayer University of Osijek |
Abstract
More than 90% of global trade transportation operates by sea, and over 15% is done using containers. Consequently, container transportation plays a significant role in global trade. Containers are stored in container yards while waiting to be loaded onto ships. Container yards have limited capacity, so containers are usually stored in stacks on each other. In most cases, the sequence in which these containers must be loaded onto the ship is unknown, making it impossible to rearrange them so that each container can be retrieved without relocating another container blocking it. The Container Relocation Problem (CRP) is a combinatorial optimization problem tasked with finding a sequence of container relocations to retrieve all containers in a defined order while optimizing one or more criteria. CRP is an NP-hard problem, meaning that, in most cases, it can not be solved exactly. Because of that, heuristic approaches are generally used to solve it. The simplest heuristic methods for solving CRP are Relocation Rules (RRs). RRs are fast and simple, but their creation requires domain knowledge and must be developed for each criterion separately. Within this study, the process of developing RRs will be automated using Genetic Programming (GP) to overcome this problem. Experimental results show that RRs developed using GP achieve better results than existing manually developed rules and have good generalization ability, which makes this approach a viable option.
Keywords
- Nature inspired methods and algorithms
- Optimization in industry, business and finance
Status: accepted
Back to the list of papers