Knowledge acquisition and distributed decision making

A concept of knowledge acquisition and distributed decision-making for autonomous distributed robotic systems is presented. The concept is based on a distributed genetic algorithm and a supplement operation, proposed to supply a deficiency of knowledge for decision making. Considering the localization of knowledge and communication in a distributed robotic system, three issues are developed, knowledge acquisition, cooperative problem solving, and coherence of solutions. As an example of cooperative problem solving, path planning for the structure configuration of the Cellular Robotic System (CEBOT) is considered. The simulation results of the influence of local knowledge and local communication in each robot are presented