Principles of Organization and Operation of an Unmanned Robotic Production with a Decentralized Controller

This work is devoted to the topical problem of elaboration of new principles of a decentralized cooperation of the computer systems, which will create a new class of unmanned robotic productions (URP). In the shortest possible time they could produce a variety of prototypes and individual items meeting the remote customers' requirements. Use of the principles of decentralized multiagent scheduling of URP ensures: firstly, high availability of the system, because it lacks the 'bottleneck" in the form of a central controller, and a failure of any of the agents does not lead to the catastrophic consequences for the entire URP; secondly, a possibility of virtually unlimited increase of the number of nodes as a part of URP because of an easy connectivity to the information communication channel; and, finally, thirdly, a reduction of the computational load on an individual software agent in the solution of the scheduling tasks, which in turn, makes it possible to get solutions in real time. This article describes the generic principles of organization and functioning of the unmanned robotic productions with a decentralized controller. Implementation of these principles ensures: - a possibility of automation of the approach to the flow of tasks of production of various products coming from customers in advance and in unknown times; - an adaptive quasi-optimal allocation of resources based on their expertise in the real-time of receipt of the assignments; - a high payload operation of the nodes in the URP; - a possibility of unlimited scalability; - an increased fault tolerance, since it lacks components, the failure of which could lead to a failure of the entire system.

безлюдное роботизированное производство, автоматический диспетчер, граф задания, программный агент, коллективное взаимодействие, алгоритм работы агента, unmanned robotic manufacturing, decentralized controller, fault tolerance, multi-agent system, robotic machining center

1. URL: www.umpro.ru

2. URL: www.rosbalt.ru

3. Каляев А. И., Каляев И. А. Метод децентрализованного управления распределенной системой при выполнении потока заданий // Мехатроника, автоматизация, управление. 2015. № 9. C. 585-598.

4. Каляев А. И., Каляев И. А., Коровин Я. С. Метод муль-тиагентного диспетчирования ресурсов в гетерогенной облачной среде при выполнении потока задач // Вестник компьютерных и информационных технологий. 2015. № 11. C. 34-40.

5. Kalyaev A. I. Multiagent Approach for Building Distributed Adaptive Computing System // Procedia Computer Science. 2013. Vol. 18. P. 2193-2202 (URL: http://dx.doi.org/10.1016/j.procs. 2013.05.390)

6. Kalyaev A. I., Korovin Y. S. Adaptive Multiagent Organization of the Distributed Computations // AASRI Procedia. 2014. Vol. 6. P. 49-58 (URL: http://dx.doi.org/10.1016/j.aasri.2014.05.008).

7. Рейнгольд Э., Нивергельт Ю., Део Н. Комбинаторные алгоритмы. Теория и практика. М.: Мир, 1980. 476 с.