Multi-Hall Sensor Fusion for AC Signal Reconstruction with Dynamic Mode Decomposition

To address the limitations of traditional current measurement methods in non-contact measurement, high-frequency response, and anti-interference capabilities, an AC signal reconstruction and analysis technology based on multi-Hall sensor fusion and Dynamic Mode Decomposition (DMD) is proposed. A multi-Hall sensor array is designed, where multiple Hall sensors are positioned at different locations to monitor the same signal. The DMD method is employed to decompose the AC signals collected by the sensors, and an optimal truncation rank is selected for decomposition. The quality of the reconstructed signal is assessed using the root mean square error (RMSE) and signal-to-noise ratio (SNR). Subsequently, a weighted synthesis approach integrates information from all sensors, where the weights consider the distance of each sensor from the conductor and the RMSE of individual sensors, leading to a more accurate reconstructed signal. This method is applied to complex current scenarios in power systems, achieving satisfactory reconstruction performance.