Neural network guided robot collectivity: an experimental setup

Abstract

The paper presents an experimental for a mobile robot collectivity operating in a common 2- dimensional environment with unknown, stationary, obstacles. The main feature of the proposed solution is that the most important tasks in the robot collectivity (cooperation between robots, sensorial data processing and data fusion, global and local path planning, etc.), are implemented using artificial neural network (ANN) and, in particularly, cellular neural networks (CNN). The structure of the mobile robots has been reduced to a minimum, in order to increase the number of members of the collectivity and, in this way, to gain the complexity of the possible experiments.