부산대학교 도서관

This paper presents a new technique for monitoring and diagnosing faults in power transformers using a Wireless Sensor Network (WSN). The study addresses the necessity for an effective and dependable system to detect and diagnose transformer faults in real time, facilitating proactive maintenance and reducing costly downtime. The proposed methodology involves deploying wireless sensors on the transformer to gather data on temperature, oil level, vibration, and electrical parameters. This data is transmitted wirelessly to a central monitoring unit for analysis, where advanced signal processing techniques and machine learning algorithms are used to detect and diagnose transformer faults based on sensor data. Experimental tests were conducted to validate the effectiveness of the proposed system, demonstrating its ability to successfully detect and diagnose various types of transformer faults with high accuracy and reliability while providing real-time alerts for proactive maintenance actions. The significance of this study lies in its contribution to enhancing power distribution systems' reliability and efficiency by enabling early fault detection and diagnosis in transformers. The proposed WSN-based monitoring system offers a practical solution that can be implemented in existing power grids without significant infrastructure modifications.