A Wireless Weatherproof Acoustic Sensor System to Detect Anomalies in Substation Power Transformers

Power transformers are essential components in electrical substations. Depending on the operating conditions, they can have failures capable of damaging the functioning of the power grid. To mitigate these problems, it is suitable for the transmission and distribution companies to implement a sensor system capable of monitoring the equipment and detecting anomalies in its operation. Acoustic measurements and signal processing present a promising non-invasive method for monitoring the working condition of power transformers. To achieve this goal, the authors developed an acoustic sensor system, featuring a suitable microphone, battery power supply, weatherproof enclosure, signal processing unit, and Wi-Fi connectivity. Additionally, they proposed an audio processing algorithm and a machine learning model for anomaly detection that analyzes the audio parameters and provides the transformer diagnosis. The proposed sensor system with anomaly detection approach can aid in preventing power transformer failures by analyzing audio parameters and providing valuable insights into the equipment's health. With this information, maintenance teams can intervene before critical failures occur, avoiding financial losses and interruptions in the power supply. The proposed approach can be readily integrated into existing substation infrastructures, enhancing the overall reliability of the power grid.