Pathfinding with Obstacle Avoidance using a Genetic Algorithm
Master thesis
Permanent lenke
http://hdl.handle.net/11250/260929Utgivelsesdato
2013Metadata
Vis full innførselSamlinger
Sammendrag
In this paper a trajectory planner for a 3D obstacle avoidance problem was researched and implemented. The resulting choice was based on results gathered from implementing a Genetic algorithm on a simplified 2D problem in the project leading up to, and included in, this paper. The use of a Genetic Algorithms was chosen on the merits of good adaptability to a wide range of problems, solution effectiveness and solving speed for complex problems. The 2D problem tested the effectiveness of a Genetic Algorithm in continuous space and its Pathfinding abilities. The tested 2D algorithm was expanded to a 3D environment and implemented on a specific robot manipulator, the Universal Robot 5. The generated paths are presented and the main design principles introduced and discussed.