Planning the Behavior of an Autonomous Flying Robot in a Space of Subtasks. Knowledge Representation Model

Abstract This article shows that usually the automatic control system of autonomous flying robots based on unmanned aerial vehicles has limited computing resources, which makes it impossible to use known labor-intensive logical models of knowledge representation and processing to plan goal-directed behavior. Thus, there is a need to develop a model of knowledge representation and processing that makes it possible to plan goal-oriented behavior under conditions of a priori uncertainty of the problem environment with polynomial complexity. To solve this problem, a model of knowledge representation is constructed in the form of a set of typical basic, intermediate, and terminal growth elements used to automatically plan goal-seeking behavior in the space of subtasks in the form of a growing reduction network model for solving complex problems under uncertainty. Automatic goal-setting procedures are developed that allow an autonomous flying robot to secure its activities under a priori uncertainty in an unstable problem environment.