Intellectual Property India Publishes Patent Application for Ai-Enabled Federated Digital-Twin System For Predictive Fault Diagnosis, Cyber-Attack Detection, And Self-Healing Control Of Distributed Power-Electronic Networks

MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202621095990 A) filed by Vikramsingh Ravindrasingh Parihar; Mehatab Mahibub Sanadi; Aditya Arvind Ramteke; Anuj Bhimaji Kandalkar; Sakshi Santosh Pawar; Swati Kishor Mohite; Shaikh Saminabano Ayyub; Geetika Narang; Snehal Y. Kulkarni; and Harshada M. Raghuwanshi on August 08, 2026, for Ai-Enabled Federated Digital-Twin System For Predictive Fault Diagnosis, Cyber-Attack Detection, And Self-Healing Control Of Distributed Power-Electronic Networks.

Inventors include Mehatab Mahibub Sanadi; Aditya Arvind Ramteke; Anuj Bhimaji Kandalkar; Sakshi Santosh Pawar; Swati Kishor Mohite; Shaikh Saminabano Ayyub; Geetika Narang; Snehal Y. Kulkarni; Harshada M. Raghuwanshi; and Vikramsingh Ravindrasingh Parihar.

The application for the patent was published on September 25, 2026, under issue no. 39/2026.

Abstract: An artificial-intelligence-enabled federated digital-twin system for a distributed power-electronic network comprises converter nodes having sensors, local controllers, edge processors, and physical switching interfaces. Each edge processor executes a node-level electrical or electrothermal digital twin and generates residuals between measured and predicted behaviour. A predictive engine determines component degradation, fault class, severity, or remaining useful life. A cyber-attack detector evaluates physical measurements, converter commands, communication metadata, topology, and cross-node consistency to detect false-data injection, replay, command tampering, unauthorized switching, or model poisoning. Authenticated model updates are aggregated through trust-weighted federated learning while raw operational data remains local. A topology-aware graph twin localizes affected nodes and predicts network consequences. A self-healing coordinator generates derating, control- mode, bypass, rerouting, isolation, quarantine, or load-management actions. A safety supervisor verifies electrical, thermal, stability, protection, and authorization constraints before physical execution, while an independent fallback controller maintains safe operation when artificial-intelligence processing is unavailable.

Disclaimer: Curated by HT Syndication.