System of allocation and planning for the performance of tasks by agents of swimming robot systems on the basis of collective decision making

Abstract

The introduction of logistics robots in automated warehouses provides a significant increase in quality and a reduction in logistics costs. Recently, more and more attention has been paid to the decentralization of the management of logistics robots, which ensures the flexibility, reliability and scalability of the system. Based on this, logistics robots operating in an automated warehouse are often considered as swarm robotic systems (SRS). Known methods for distributing and scheduling tasks used as part of software and algorithmic support for centralized control systems of logistics robots do not take into account the limitations of the sensory and computing capabilities of robotic devices used as part of the SRS (for example, a small amount of RAM, low clock frequency of the processor, low battery capacity, low performance of on-board sensors and sensors, etc.). This fact indicates the absence of a comprehensive solution that allows implementing as a swarm robotic system. Thus, the present work is aimed at developing an approach to the synthesis of a decentralized system for distributing and planning the execution of tasks by agents of swarm robotic systems when operating in an automated warehouse. A distinctive feature of the proposed approach is the use of a collective decision-making procedure when distributing tasks between agents, which ensures the absence of collisions (choosing the same task by two or more agents). An element of scientific novelty in the framework of the proposed approach is the method of collective decision-making, which is distinguished by a modified procedure for choosing the best alternative in accordance with the characteristics of the task distribution process in swarm robotic systems. The reliability of the obtained results was confirmed in the course of simulation modeling using the simulator of multi-agent robotic systems ARGoS. The presented approach to the synthesis of a system for distributing and scheduling the execution of tasks based on collective decision making can be used in modeling decentralized control systems for swarm robotic systems.