Simulation of Autonomous Agents Collective Behavior According to the Consolidation Scenario

Abstract

Purpose. To perform research on group behavior in a multi-agent system during performing a consolidation task by simulation modeling in the created by the bionic prototype modeling environment. The research was considered as a preliminary step towards the creation of applied systems of group robotics. Methodology. We have applied graph theory, mathematical differential calculation, simulation modeling. Results. We have developed a mathematical model of the problem of consolidation of agents in a multi-agent system according to the scenario of leader persecution. The software implementation of the environment for studying the behavior of multi-agent systems in real time has been performed. Simulation of a multi-agent system with variation of system parameters and experimental conditions was performed. The directions of improvement of the modeling environment in order to bring the simulation closer to the real conditions of functioning of applied multi - agent systems are determined. Originality. We have developed an original research software environment for real-time swarm robotic tasks. Forms and means of monitoring the state of the multi-agent system and individual agents, suitable for rapid analysis by the researcher of the model experiment, are proposed. We have experimentally established the effectiveness of the proposed approach in the process of modeling the consolidation of a multicomponent system with high dynamic parameters. Practical value. We have tested the effectiveness of the bionic model of behavior of a multi-agent system in solving the problem of consolidation in the scenario of convergence of agents to the leader. Сonclusions. Practical usage of received results is possible at all stages of the life cycle of the swarm robotic applying intellectual systems, in particular, during identifying the optimal structure of links in the group of agents and the assessment of the adequacy of their behavior in the functional space. A further development of the research environment is in expanding functional possibilities to describe the scenes, the properties of the agents, the communication systems in the group, the types of consolidation. References 17, tables 0, figures 5.