Adaptive Management of Multi-Agent Intelligent System: Algorithmic Aspect

In the process of constructing a multi-agent intelligent system, the stages of structural and functional decomposition of the system and the determination of its tasks, the allocation or formation of the necessary groups of agents, implementation of system administration procedures and operational control of the operability of all components of the system are considered, as a rule. In this regard, a comparison was carried out of various options for the structural construction of a multi-agent system, taking into account the flexibility of control, the possibilities for functional redundancy of its components and their reconfiguration, which made it possible to identify and recommend a tree-like network topology for widespread application. Objects of planning and monitoring the effectiveness of the actions of executive agents, as well as an object of planning and monitoring the effectiveness of the system's actions, have been introduced as nodes of the topology. It is shown that the organization of adaptive management requires a formalized representation in the objects and agents of the system of certain segments of the system topology, as well as the states of the external environment, planning authorities, system agents, channels for ensuring information and technical interaction and functional tasks. A rational method of such mapping is the construction by systems analysts and systems engineers of system composition logs, composition of topology segments, and agent operation logs in the form of logical-multiple relations. The specified relationships link various types of objects, agents, and tasks with attributes of system characteristics, parameters, and states of system operability. A fundamental possibility of determining a rational composition of relation domains that allows for functional expansion is presented. The algorithm for functioning of a multi-agent system is described, in which each cycle is based on the results of operational scanning of individual logs of the composition of agents and their functioning, as well as selection from the relations of tuples of the next tasks to ensure the subsequent solution of a specific instance of the task. With the help of special alerts in the circular transmission mode, self-synchronization and adaptive selection of work by system agents within each segment of the topology are ensured.

1. Gulay A. V., Zaytsev V. M. (2020) Convergence of Intelligent Systems. Minsk, Publishing House of Information and Computing Center of the Ministry of Finance. 384 (in Russian).

2. Zaytsev V. M. (1982) Organization of Distributed Data Processing on Automated Control Systems. Voprosy Radio-elektroniki. Ser. Obshchetekhnicheskaya [Questions of Radio Electronics. General Technical Series], Iss. 10, 26–32 (in Russian).

3. Gorodetsky V. I., Grushinsky M. S., Khabalov A. V. (1998) Multi-Agent Systems (Review). Novosti Iskusstvennogo Intellekta [Artificial Intelligence News], (2), 64–116 (in Russian).

4. Karpov V. E. (2016) Models of Social Behavior in Group Robotics. Upravlenie Bolshimi Sistemami [Managing Large Systems], Iss. 59, 165–232 (in Russian).

5. Tarasov V. B. (2002) From Multi-Agent Systems to Intelligent Organizations. Moscow, Editorial URSS Publ. 352 (in Russian).

6. Guessoum Z., Briot J.-P., Faci N., Martin O. (2010) Towards Reliable Multi-Agent System: An Adaptive Replication Mechanism. Multiagent and Grid Systems, 6 (1), 1–24. https://doi.org/10.3233/mgs-2010-0139.

7. Hübner J. F., Boissier O., Bordini R. H. (2011) A Normative Programming Language for Multi-Agent Organizations. Annals of Mathematics and Artificial Intelligence, 62 (1), 27–53. https://doi.org/10.1007/s10472-011-9251-0.

8. Boissier O., Bordini R. H., Hübner J. F., Ricci A. (2019) Dimensions in Programming Multi-Agent Systems. Knowledge Engineering Review, 34 (2), 1–28. https://doi.org/10.1017/ s0 26988891800005x.

9. Erofeeva V. A., Ivansky Yu. V., Kiyaev V. I. (2015) Control of A Swarm of Dynamic Objects Based on A Multi-Agent Approach. Kompyuternye Instrumenty v Obrazovanii = Computer Tools in Education Journal, (6), 34–42 (in Russian).

10. Vorobyov V. V. (2017) Leader Selection and Clusterization Algorithms in a Static Robot Swarm. Mekhatronika, Avtomatizatsiya, Upravlenie, 18 (3), 166–173. https://doi.org/10.17587/mau.18.166-173 (in Russian).

11. Zaytsev V. M. (1981) The Phenomenon of Interference in Automated Control Systems Telecode Networks. Voprosy Radioelektroniki. Ser. obshchetekhnicheskaya [Questions of Radio Electronics. General Technical Series], Iss. 12, 35–41 (in Russian).

12. Zaytsev V. M. (1981) Determination of Data Transmission Parameters Based on the Principle of Information Stabilization. Voprosy Radioelektroniki. Ser. Obshchetekhnicheskaya [Questions of Radio Electronics. General Technical Series], Iss. 12, 42–51 (in Russian).

13. Gulay A. V., Zaitsev V. M. (2016) Transmission Reliability of Transactions in Mechatronic Systems: Choice of Triplets for Noiseproof Code. Mekhatronika, Avtomatizatsiya, Upravlenie, 17 (1), 26–31. https://doi.org/10.17587/ mau.17.26-31 (in Russian).

14. Gulay A. V., Zaitsev V. M. (2023) Reed-Solomon Error-Correcting Codes: Application in Intelligent Systems Technology. Current Issues in the Development of Modern Science and Technology. Petrozavodsk, Publishing House of International Center for Scientific Partnership “Novaya Nauka”, 377–406 (in Russian).

15. Smith S. (2008) Digital Signal Processing. Moscow, Dod-Eka-XXI Publ. 718 (in Russian).