COGNITIVE ARCHITECTURE FOR ROBUST ADAPTIVE-CONTROL OF ROBOTS IN A TEAM

The objective of this paper is to present a cognitive architecture tha t utilizes three different methodologies for adaptive, robust control of robots behaving intelligently in a team. The robots interact within a world of objects, and obstacles, performing tasks robustly, while i mproving their performance through learning. The adaptive control of t he robots has been achieved by a novel control system. The Tropism-bas ed cognitive architecture for the individual behavior of robots in a c olony is demonstrated through experimental investigation of the robot colony. This architecture is based on representation of the likes and dislikes of the robots. It is shown that the novel architecture is not only robust, but also provides the robots with intelligent adaptive b ehavior. This objective is achieved by utilization of three different techniques of neural networks, machine learning, and genetic algorithm s. Each of these methodologies are applied to the tropism architecture , resulting in improvements in the task performance of the robot team, demonstrating the adaptability and robustness of the proposed control system.