Multi-agent Planning of Autonomous Robots Based on Declarative Rules
Aleksandr Ermakov, Larisa Suchkova · 2020 International Multi-Conference on Industrial Engineering and Modern Technologies (FarEastCon) · 2020
In this work, we consider the rules declaratively describing the goal states of environment. A multi-agent autonomous system based on such rules has the ability to distribute subtasks between agents and complete its goals. There is a need to create a set of rules that would fully describe the behavior of both individual agents and the behavior of a multi-agent system. The work describes syntax of the rules and system architecture of an autonomous robot. We have implemented a rule parser, a dependency network, and an interpreter. The interpreter was integrated with applied software that implements search algorithms and provides decision trees for decision making. In the process of global planning, the search for the optimal solution is carried out, taking into account time spent and resources. Planning is a continuous action that allows you to react to unexpected situations and to correct inaccuracies in the modeling. Finally, we wrote declarative rules and carried out an experimental test of the rules processing system in a virtual environment.