Интеллектуальная система поддержки обучения специалистов радиорелейной связи на основе архитектуры RAG

Abstract

The article considers the problem of improving the quality of training of radio-relay communication specialists through the introduction of an intelligent assistant based on the RAG (Retrieval-Augmented Generation) architecture. The developed system «AXON» provides semantic search and generation of reliable answers only on the basis of a verified knowledge base, operates in a local secure loop without access to external networks. The structure, algorithms and test results of the system are described. The ap plication of RAG technology made it possible to combine the flexibility of a dialogue interface with strict requirements for reliability and information security in military education.