In this paper a study of termination functions (stop criterion) for evolutionary path planning algorithm is presented. Tested algorithm is used to determine close to optimal ship paths in collision avoidance situation. For this purpose a path planning problem is defined. A specific structure of the individual path and fitness function is presented. For the simulation purposes a close to real tested environment is created. Five independent tests are performed using three stopping conditions: termination upon number of generations, population average score and population diversity. The results are compared in terms of number of generations needed to stop algorithm, calculation time and fitness function value. Undertaken research allow to select suitable parameters for studied stop criteria and conclude whether the termination function used in tests are adequate for specialized path planning evolutionary algorithm.
Authors
Additional information
- DOI
- Digital Object Identifier link open in new tab 10.1109/mmar.2014.6957428
- Category
- Aktywność konferencyjna
- Type
- materiały konferencyjne indeksowane w Web of Science
- Language
- angielski
- Publication year
- 2014
