An Intelligent Predictive Maintenance Framework for High Voltage Circuit Breakers in Smart Grids Using Condition Monitoring Data

Circuit breakers play a crucial role in power system protection, ensuring the safety and reliability of electrical networks. However, their performance can degrade over time due to mechanical wear, environmental conditions, and operational stress. Early detection is essential to prevent protection system failures. To this end, this paper presents a novel online monitoring framework for assessing the health condition of circuit breakers in real-time, enabling predictive maintenance and improving system reliability. The proposed approach utilizes advanced signal processing, and data analytics to analyze circuit breaker operation parameters such as closing and opening times, coil current, and transient signals. Then, it uses those data to train a novel Kernal density Isolation Forest based algorithm for status identification of on grid circuit breakers. Case studies and experimental results demonstrate the effectiveness of the proposed method in identifying early signs of circuit breakers degradation and preventing unexpected failures.