Monitoring of an Electric Motor through Linear Kalman Filter based Fusion Operation on Delayed Sensors

This study investigates a Linear Kalman Filter (LKF)-based sensor fusion framework for the monitoring of electric motors. Current and vibration signals were collected from a laboratory-scale test bench and fused using a calibrated LKF algorithm, which aligns cutoff frequency with critical spectral indicators to minimize information loss. Controlled delays were artificially introduced into one of the measurement channels to evaluate the robustness of the fusion scheme. Performance was assessed using correlation, coherence, peak error, and PSD-based metrics. Results show that the fused signal preserves the characteristic frequencies of the original measurements while exhibiting sensitivity to phase misalignment. Although increasing delays caused fluctuating time-domain similarity and amplitude errors, the frequency-domain content remained robustly conserved. These findings highlight the robustness and reliability of LKF-based fusion against moderate sensor delays and its potential as a reliable tool for electric motor monitoring.