A Computer Simulation of Braitenberg Vehicles
AIR FORCE INST OF TECH WRIGHT-PATTERSON AFB OH SCHOOL OF ENGINEERING
Pagination or Media Count:
In his treatise, Vehicles Experiments in Synthetic Psychology, Valentino Braitenberg used simple, autonomous vehicle designs to illustrate the principles behind animal nervous system organization and operation. The goal of this effort was to produce a computer program to allow a researcher to experiment with these concepts and to analyze their performance using methods employed by experimental psychologists. The resulting program allows the user to design vehicles that respond to changes in their environments and that have the ability to adapt their behavior, using a learning algorithm developed by Teuvo Kohonen. The vehicle designer is free to select sensor attributes, numbers of neurons, learning periods, environments, and starting knowledge. The vehicle may be analyzed by tracking their success at locating food in their environment. The food is located close to stimuli to which the vehicle sensors can respond. Food discovery triggers the learning algorithm, adapting behavior to improve food finding-performance. The initial evaluations failed to provide convincing proof that the simple vehicles tested had succeeded in totally adapting to their environments.
- Land and Riverine Navigation and Guidance