A cognitive digital twin architecture for intelligent gas distribution systems

This is the authors' abstract. We don't add key points for this paper.

Gas Distribution Systems (GDS) are critical infrastructures for industrial facilities and large buildings because they provide continuous gas delivery while requiring strict safety, reliability, and monitoring measures. This study presents a smart GDS architecture that integrates cognitive digital twins with Software Defined Networking (SDN), Internet of Things (IoT), deep learning, and Virtual Reality (VR). The main novelty is the design of a unified cognitive digital twin framework that connects communication management, leak detection, and immersive operator interaction in one architecture. The proposed system uses SDN-assisted LoRaWAN communication for adaptive data forwarding, an RNN-LSTM model supported by statistical anomaly analysis for gas leak detection, and a VR interface for situational awareness, remote inspection, and user control. LoRaSim-based evaluation indicates improved packet delivery and reduced network energy consumption compared with a single-hop baseline. Leak detection results show high performance for most sensors, while the third sensor remains more difficult because of stronger flow variability. The VR assessment also reports positive operator perception of satisfaction, applicability, and effectiveness. These results show that combining AI-driven analytics, adaptive communication, and immersive visualization can improve the resilience, operational intelligence, and decision support capabilities of smart GDS infrastructures.