Abstract:
Challenges and complex problems arising in the resource management of modern enterprises are considered. The existing resource planning models, methods and tools for enterprises are reviewed, and new requirements for adaptive multicriteria resource planning in real time are presented. The concept of autonomous artificial intelligence (AI) systems for adaptive resource planning based on multi-agent technologies is discussed. The evolution of the approach to solving complex resource management problems is described: from traditional optimization of a single objective function, ignoring the individual interests of participants, to game theory with their competition and cooperation. The approach to finding and maintaining a competitive equilibrium (consensus) between participants is further developed via conflict identification and negotiations for conflict resolution with mutual trade-offs. A basic model of a multi-agent demand-supply network with a virtual market and a compensation method for reaching consensus for adaptive resource planning are presented. The functionality and architecture of intelligent adaptive resource planning systems are considered. The implementation results of AI solutions for industrial applications are provided, and the possibility of improving the effectiveness of resource usage by enterprises is shown. Finally, the lessons learned from the experience in R&D work and the prospects of this approach are discussed.
Fulltext:
PDF file (1353 kB)