MUMBAI, India, July 30 -- Intellectual Property India has published a patent application (202641087405 A) filed by Dr. N. Satheesh; Dr. Manikandan Parasuraman; Dr. Renukadevi P; and Dr. B. Swamynathan on July 17, 2026, for Machine Learning–enhanced Fault Detection Module For High-Voltage Electrical Protection Systems.
Inventors include Dr. N. Satheesh; Dr. Manikandan Parasuraman; Dr. Renukadevi P; and Dr. B. Swamynathan.
The application for the patent was published on July 24, 2026, under issue no. 30/2026.
Abstract: ABSTRACT Machine Learning–Enhanced Fault Detection Module for High-Voltage Electrical Protection Systems The proposed invention, titled “Machine Learning– Enhanced Fault Detection Module for High-Voltage Electrical Protection Systems,” presents an intelligent and adaptive approach to improving the reliability, speed, and accuracy of fault detection in modern electrical power networks. Conventional protection systems primarily rely on fixed threshold-based relays and deterministic logic, which often struggle to accurately identify complex fault conditions, transient disturbances, and evolving grid dynamics, especially in the presence of renewable energy integration and fluctuating load patterns. To overcome these limitations, the proposed system integrates advanced machine learning techniques with real-time electrical signal monitoring to create a robust and data-driven fault detection framework. The module continuously acquires high-resolution data such as voltage, current, frequency, and phase angle from sensors and intelligent electronic devices deployed across high-voltage transmission and distribution systems. This data is then subjected to preprocessing steps including noise filtering, normalization, and feature extraction using advanced signal processing methods such as wavelet transforms and Fourier analysis to capture both steady-state and transient characteristics of the system. The processed data is fed into a machine learning engine trained on a comprehensive dataset that includes normal operating conditions, multiple fault types such as line-to-ground, line-to-line, and three-phase faults, as well as non-fault disturbances like switching operations and power swings. The system employs a hybrid learning architecture combining deep neural networks and anomaly detection algorithms to accurately classify faults and distinguish them from benign disturbances, thereby reducing false alarms and unnecessary tripping. A key feature of the invention is its adaptive learning capability, which allows the model to continuously update and improve its performance based on newly acquired data, ensuring long- term effectiveness even as grid conditions change. Additionally, the module includes a real-time decision-making unit capable of triggering protective actions such as circuit breaker operation and fault isolation within milliseconds, minimizing equipment damage and ensuring system stability. The system is designed for seamless integration with existing SCADA and smart grid infrastructures, enabling easy deployment without significant hardware modifications. Furthermore, it incorporates a user-friendly visualization dashboard that provides operators with actionable insights, including fault classification results, system health indicators, and predictive maintenance recommendations. By leveraging machine learning and real-time analytics, the proposed invention significantly enhances the efficiency, resilience, and intelligence of high-voltage electrical protection systems, making it highly suitable for next-generation smart grids and modern power networks.
Disclaimer: Curated by HT Syndication.
