Agent-based Intelligent Decision Support System: Architecture and development

The article considers the architecture, analysis, design and implementation of the Agent-based Intelligent Decision Support System. It describes the structure of software agents, the organization of their joint work, the functions implemented by software agents and queries to the distributed knowledge base, and algorithms for multi-agent reinforcement learning. The authors consider methods for increasing the system’s performance and reliability, algorithms for optimizing the logical structure of the knowledge base. A subsystem for managing a distributed knowledge base is described, capable of adapting to changing environmental conditions, as well as responding in a timely and adequate manner to situations arising from a shortage of computing or time resources. The obtained results demonstrate the effectiveness of the agent-oriented approach for creating an intelligent decision support system capable of finding and analyzing solutions to complex problems in conditions of uncertainty and a dynamic environment.