MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202621093165 A) filed by Vikramsingh Ravindrasingh Parihar; Shlok Santosh Chavan; Ashwini Atit Kadam; Ketan Sanjay Jadhav; Sujit Ajitkumar Ahirekar; Soham Mahendra Thoke; and Viraj Anil Barge on July 31, 2026, for Ai-Enabled Multi-Modal Dc Arc-Fault Protection System With Converter-Embedded Diagnostic Signal Injection And Selective Solid-State Isolation For Renewable-Energy Microgrids.
Inventors include Shlok Santosh Chavan; Ashwini Atit Kadam; Ketan Sanjay Jadhav; Sujit Ajitkumar Ahirekar; Soham Mahendra Thoke; Viraj Anil Barge; and Vikramsingh Ravindrasingh Parihar.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: An artificial-intelligence-enabled protection system for a direct-current renewable-energy microgrid comprises multi-modal branch sensors, an electronically controlled power converter, a machine-learning engine, and selectively controllable solid-state isolation switches. Passive current, voltage, high-frequency, temperature, and optional physical-sensor features identify a suspected arc event. A diagnostic controller selects a bounded coded excitation according to bus, converter, load, thermal, and power-quality conditions and causes an existing converter to superimpose the excitation on normal power conversion. Synchronized branch responses are processed to determine impedance, attenuation, phase, delay, nonlinear, or baseline-residual features. The machine-learning engine fuses passive features, active-response features, topology information, sensor quality, and uncertainty to classify and locate the arc fault. A protection coordinator isolates the identified branch and verifies current interruption, while a service-continuity controller supports the bus and maintains unaffected branches. Independent fallback protection remains operable if learned classification, communication, or selective isolation fails.
Disclaimer: Curated by HT Syndication.
