Evolving Subtasks in Predator-Evader Problem

The creation of strategies to meet abstract goals is an important behavior exhibited by natural organisms. A situation requiring the development of such strategies is the predator-evader problem. To study this problem, Khepera robots are chosen as agents. Using computer simulations the evolution of the adaptive behavior is studied. Neural network architecture with evolvable weights is used as decision mechanism of the predator. Evolutionary programming is employed to evolve the predator for developing adaptive behavior to accomplish the task of catching prey. Then, this task is divided into two subtasks and the predator is evolved to accomplish these subtasks and the neural networks accomplishing subtasks are combined to form the predator. The results of these two approaches are compared