Abstract
In this paper we deal with the problem of designing preemptive patrolling itineraries at sea. As a case study we use real data provided by the Portuguese Navy, concerning search and rescue missions. This can be seen as a curve fitting problem in which we aim at predicting future occurrences based on previous reported ones. We propose a method for the design of itineraries based on the use of a 1 dimensional self-organizing map. A number of tests are performed using self-organizing maps of different size. We conclude that the proposed method has a number of relevant characteristics such as robustness, speed and flexibility which are particularly adequate to the problem in hands.
Original language | English |
---|---|
Title of host publication | Proceedings 2005 - The 8th AGILE International Conference on Geographic Information Science, AGILE 2005 |
Pages | 1-8 |
Number of pages | 8 |
Publication status | Published - 2005 |
Event | 8th AGILE International Conference on Geographic Information Science, AGILE 2005 - Estoril, Portugal Duration: 26 May 2005 → 28 May 2005 |
Conference
Conference | 8th AGILE International Conference on Geographic Information Science, AGILE 2005 |
---|---|
Country/Territory | Portugal |
City | Estoril |
Period | 26/05/05 → 28/05/05 |
Keywords
- Prediction
- Route optimization
- Route planning
- Traveling salesman problem