FMCW Radar Vibration Measurement Using Sub-Chirp Group Multicircle Fitting With PCA-Based Solution Space Constraints

This is the authors' abstract. We don't add key points for this paper.

Vibration measurements are pivotal across diverse domains, including industrial environments, medical care, and civil engineering. Non-contact measurement methods offer distinct advantages over their contact counterparts by circumventing equipment intrusion and minimizing interference with the target. This preservation of the target’s movement process from external disturbances ensures measurement accuracy and stability. However, existing non-contact microdisplacement measurement technologies grapple with a trade-off between accuracy and efficiency. To address this challenge, our study leverages linear frequency-modulated continuous wave (FMCW) radar to obtain insights from sub-chirp group investigations and analyzes the impact of extending the arc in circle fitting. Our method, termed sub-chirp group multicircle fitting (SCGMCF), estimates the circles associated with sub-chirps in a comprehensive manner. This modification significantly improves the accuracy and efficiency of amplitude measurement for microdisplacement vibration. In our method, we also propose a principal component analysis (PCA)-based method to constrain the solution space within the I/Q domain. By considering aspects of direction, amplitude, and noise, this approach enhances static component elimination and mitigates phase collapse. Furthermore, experimental results demonstrate the effectiveness of our methods, revealing successful extraction of movement processes under low vibration amplitudes of 30μm, with runtime reduced by more than 90%.