Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.

  

On the Robustness of Swarm Behavior to Obstacle Variations

Back
Abstract:
In a very general sense, the term swarm is usually meant to signify any collection of objects (agents) that interact with one another. It has long been recognized that interactive cooperative behavior within biological groups or swarms is advantageous, for example, in avoiding predators or, vice versa, in capturing prey. The way nature evolved for the coor- dination of a °ock of birds or a school of ¯sh, is a great example for controlling a group of unmanned vehicles or mobile robots. According to principles observed in nature, a swarm is modeled as a cloud of particles, where their interaction is constructed from attractive and repulsive forces, dependent on the relative distance between the individuals. For autonomous vehicles or robots, the interaction \forces" are due to motorized propulsion arising from inter-vehicle communication. The swarm is designed to be capable of navigating through a given arrangement of obstacles and reaching a target. Driven by a rapid simulation scheme, the interaction laws of the swarm are optimized for a given obstacle arrangement such that time and e®ort required to reach the target are minimized. For this purpose, a genetic algorithm is employed to handle the associated objective function, which depends in a non-convex and non-di®erentiable manner on the free variables in the system. By alteration of the arrangement of obstacles and monitoring the resulting swarm behavior, the robustness of the swarm design to obstacle variations is demonstrated.

Type of Seminar:
IfA Seminar
Speaker:
Sebastian Trimpe
Date/Time:
Jan 21, 2008   16:00
Location:

ETL, K25
Contact Person:

Melanie Zeilinger
No downloadable files available.
Biographical Sketch: