فهرست نشریات علمی دسترسی آزاد ایران

Effective dispatch control of gas transmission systems (GTS) requires quick, reliable, and interpretable decision-making in conditions of uncertainty and increased responsibility. Completely autonomous decision support systems (DSS) in such critical infrastructure facilities have limited applicability because of industrial safety requirements, regulatory restrictions, and the complexity of the algorithms. In this article, we introduce a semi-automatic decision support system for gas transmission network operations based on a natural language user interface that enables effective interaction between the dispatcher and the system's computing core. The architecture of the solution is modular and includes a dialogue module, a unit for semantic processing of queries, and a core that combines mathematical models of flow distribution, algorithms for optimization, and a knowledge base. In this work, we developed and implemented a semi-automatic decision support system with a natural language interface for gas transmission network dispatching, in which unstructured dispatcher queries are automatically converted into formalized flow balance and optimization tasks through semantic analysis and automatic model parameterization using network topology. The results show that the proposed approach reduces decision-making time by more than 80% compared to the classical one, increases the accuracy of operator instructions interpretation, and improves the quality of dispatching in normal and emergency conditions.