Enhancing transmission line protection with adaptive ANN-based relay for high resistance fault diagnosis

• Published in: Electrical engineering Vol. 106; no. 6; pp. 7117 - 7132
• Main Authors: Moparthi, Janardhan Rao, Bhukya, Krishna Naick, Chinta, Durga Prasad, Biswal, Monalisa
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.12.2024; Springer Nature B.V
• Subjects: Adaptability; Adaptive systems; Algorithms; Artificial neural networks; Back propagation; Decision making; Economics and Management; Effectiveness; Electrical Engineering; Electrical Machines and Networks; Energy Policy; Engineering; Fault detection; Fault diagnosis; Fault location; Faults; Fourier transforms; High resistance; Neural networks; Original Paper; Parameter identification; Parameter modification; Parameter robustness; Power Electronics; Real time operation; Relay systems; Relaying; Signal processing; System effectiveness; System reliability; Systems stability; Transmission lines; Wavelet transforms; Fault diagnosis; High resistance fault; Intelligent relay; Transmission line protection; Artificial neural networks
• ISSN: 0948-7921; 1432-0487
• DOI: 10.1007/s00202-024-02369-w

In modern power systems, accurate and timely detection of faults is crucial for ensuring system stability and reliability. The presence of high resistance in fault path curtails current and causes conventional distance relays to malfunction. These methods often require two-end measurements for accurate assessment of fault resistance necessitates an expensive communication channel. This paper proposes an innovative approach to enhance transmission line protection through an adaptive artificial neural network (ANN)-based relay system. The relay system integrates three ANN units: the fault detection unit, fault classification unit, and fault location unit, each tailored to detect, classify, and locate faults, respectively. By utilizing single-end measurements and employing discrete Fourier transform for feature extraction, the proposed algorithm efficiently diagnoses various fault conditions, including high resistance faults. Additionally, the algorithm dynamically updates its characteristics based on the estimated fault resistance (using one cycle post-fault data and the status of each ANN unit) in real-time, ensuring adaptability to changing system conditions, especially when the fault resistance falls beyond the scope of the training data. Simulation results on a 400-kV, 50-Hz transmission system demonstrate the robustness and effectiveness of the proposed approach in accurately identifying fault events under varying fault parameters, while also accounting for arcing faults and transducer errors. The suitability of the proposed method for real-time operations has been validated using OPAL-RT digital simulator. The adaptability of the proposed method for higher order systems is verified by performing a test case on the modified WSCC 9-bus system. The results support the adaptability and effectiveness of the proposed relaying algorithm in securing the transmission line under various conditions, including high resistance faults.