Management of Network Supply System Based on Multi-Agent Technology

Problem statement: Modern supply chains are presented by large-size models. Supply network scheduling requires considerable computing resources and a lot of time. The traditional centralized supply chain management systems cannot handle the increasing complexity of the scheduling tasks in the conditions of fast-changing market situation, fluctuations in supply and demand, various disruptions and unpredictable events. The paper describes a new approach to management of the supply networks based on multi-agent technologies and ontologies. Methods: A method of a virtual "round table" is suggested for support of decision-making during supply chain scheduling. Besides, a method of adaptive dynamic supply chain scheduling depending on the events is proposed, which is based on multi-agent technology using ontology. Results: A virtual "round table" concept was developed, as well as regulations for coordination of decisions at all the stages of product supplies, and involving all the interested participants in this process. Advantages of the suggested approach are illustrated by an example of multi-criteria decision-making under condition of company department common values. Main components of the client-server architecture for the multi-agent system of the network supply management were singled out. The decision to use ontology for description of the concept knowledge necessary for the chain supply management was substantiated. The developed models and scheduling methods ensure an adaptive dynamic supply scheduling in a real-time mode in accordance with the individual orders and demand forecast for any time-period and with any details. Practical value: application of the multi-agent system for the network supply management is most efficient in the situations, when it is necessary to adapt a schedule to small, but important deviations of the input data instead of processing a new set of data. The conducted research and modeling demonstrate that a transfer to a real-time scheduling and management will make it possible to increase the companies' profits due to a quicker adaptation to the changing conditions.

цепочки поставок, мультиагентная технология, онтология, реальное время, виртуальный "'круглый" стол, адаптивность, планирование, supply chain, multi-agent technology, ontology, real-time, virtual "roundtable", adaptability, scheduling

1. Бурков В. Н. и др. Механизмы управления. Управление организацией: планирование, организация, стимулирование, контроль. М.: ЛЕНАНД, 2013. 215 с.

2. Dynamic modeling and control of supply chain systems: A review // Computers & Operations Research. 2008. Vol. 35, Iss. 11. P. 3530-3561.

3. Rzevski G., Skobelev P. Managing Complexity. Southampton, UK, WIT Press, 2014. 202 c.

4. Leung J. Handbook of Scheduling: Algorithms, Models and Performance Analysis // CRC Computer and Information Science Series, London: Chapman and Hall / CRC, 2004. 1216 p.

5. Leitao P., Vrba P. Recent Developments and Future Trends of Industrial Agents // Proc. of 5th Int. Conf. on Holonic and Multi-Agent systems in Manufacturing (HoloMAS 2011), France, Tolouse, 2011. Springer, Berlin. P. 15-28.

6. Yann Chevaleyre at al. Issues in Multiagent Resource Allocation // Informatica. 2006. Vol. 30. P. 3-31.

7. Мирошник И. В., Никифоров В. О., Фрадков А. Л. Нелинейное и адаптивное управление сложными динамическими системами. СПб.: Наука, 2000. 549 с.

8. Pinedo M. Scheduling: Theory, Algorithms, and Systems. Springer, 2008. 664 p.

9. Malti Baghel, Shikha Agrawal and Sanjay Silakari. Survey of Metaheuristic Algorithms for Combinatorial Optimization // Int. Journal of Computer Applications. 2012. Vol. 58, N. 19. P. 21-31.

10. Davis R., Burns A. A survey of hard real-time scheduling for multiprocessor systems // ACM Comput. Surv., 2011. Vol. 43, N. 4, article 35.DOI= 10.1145/1978802.1978814.

11. Barsanti L., Sodan A. Adaptive Job Scheduling Via Predictive Job Resource Allocation // Job Scheduling Strategies for Parallel Processing. Springer: Lecture Notes in Computer Science. 2007. Vol. 4376. P. 115-140.

12. Madsen В., Rzevski G., Skobelev P., Tsarev A. Real-time multi-agent forecasting & replenishment solution for LEGOs branded retail outlets // Proc. of 13th ACIS International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel / Distributed Computing (SNPD 2012), August 8-10, 2012, Kyoto, Japan. Springer, 2012. P. 451-456.