Abstract:
The main features of the organization of problem solvers for autonomous mobile intelligent agents of various purposes, capable of collectively solving complex problems associated with the need to plan purposeful activities in a priori undescribed conditions of a problem environment, are considered. Heuristic rules for inference and procedures for planning purposeful behavior have been developed, which endow autonomous mobile intelligent agents with the ability to autonomously solve subtasks assigned to them, in the process of joint implementation of a complex task posed to the team. Boundary estimates of the complexity of the proposed automatic behavioral planning procedures have been determined, which make it possible to determine the class of tasks that can be implemented on the on-board computing system of autonomous mobile intelligent systems for various purposes.
Keywords:autonomous intelligent agent, division of tasks into subtasks. collective activity, a priori undescribed problematic environment, common goal of behavior, procedures for planning behavior.