Abstract
Over the past decade, multi-agent systems (MAS) have become widespread, especially in the context of advances in smart electromechanical systems (SEMS) and the solution of distributed problem in swarm control. The properties of agents, such as autonomy and reactivity, as well as the possibility of dynamic interaction, make it possible to implement their cooperative actions in achieving common goals, and motivate SEMS a matter of serious scientific interest. Multi-agent robotic systems, combining different specialization of robot-agents, are able to accomplish assignments without any external interference, which ensure high reliability and adaptability of such systems. In order to successfully complete a common task, robot-agents must conduct complex negotiations, cooperate and coordinate their actions with each other. Each part of multi-agent robotic systems is impossible achievement common goal without the dynamic redistribution of tasks between agents in changing environmental. Purpose of research: The main approaches to the construction of models for the distribution of tasks among the MAC were analyzed. The various algorithms were compared and their mathematical modeling were established. Results: The centralized and decentralized methods of distribution tasks among agents were investigated with the aim of achieving an optimal result in minimal time and no conflicts in SEMS. Practical significance: The presented algorithms can be used to control a multi-agent system, considered as SEMS, especially to complete tasks, which are critical to the execution time, such as search and rescue operations in the case of natural or man-made disasters.
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call