MUMBAI, India, July 13 -- Intellectual Property India has published a patent application (202641081676 A) filed by Dr. Yogeesh N on July 02, 2026, for Graph Neural Network For Grid Event Propagation Inference From Media Streams.
Inventors include Dr. Yogeesh N; Dr. Shankaralingappa B M; Dr. Raghavendra H M; and Dr. Nagendra N.
The application for the patent was published on July 10, 2026, under issue no. 28/2026.
Abstract: ABSTRACT "Graph Neural Network for Grid Event Propagation Inference from Media Streams" An apparatus and method for inferring electric-grid event propagation using a graph neural network (GNN) that fuses grid telemetry with time-aligned media signals. For PMU/SCADA measurements- a grid-telemetry interface; for authenticated notices/broadcaster schedules/media intake it is a media interface; and for curated public alerts it should be an alerting system (that can channel in various high-quality sources, as clarified below). A synchronization module organizes the streams to produce a dynamic graph in which nodes are grid assets and edges encode electrical connectivity and transfer impedance with media-derived salience weights. A message-passing GNN on edge gateway or control-centre server predicts and prioritizes alerts for SCADA/ EMS/HMI via protocol adapters, inferring event origin, propagation path and arrival times; downstream impact scores. Technical Effects: Quantized, power-aware inference (minimum probability of error) provided us with reduced mean time-to-detect, enhancement in localization accuracy and reduction faux-alarm rate under real-time latency and power budgets. The invention is compatible with existing utility communication standards and can work in conjunction with centralized instances, enabling failover.
Disclaimer: Curated by HT Syndication.
