Accurate expression of neck motion signal by piezoelectric sensor data analysis

The development of high-precision sensors using flexible piezoelectric materials has the advantages of high sensitivity, high stability, good durability, and lightweight. The main problem with sensing equipment is low sensitivity, which is due to the mismatch between materials and analysis methods, resulting in the inability to effectively eliminate noise. To address this issue, we developed the denoising analysis method to motion signals captured by a flexible piezoelectric sensor fabricated from poly-L-lactic acid (PLLA) and polydimethylsiloxane (PDMS) materials. Experimental results demonstrate that this improved denoising method effectively removes noise components from neck muscle motion signals, thus obtaining high-quality, low-noise motion signal waveforms. Wavelet decomposition and reconstruction is a signal processing technique that involves decomposing a signal into different scales and frequency components using wavelets and then selectively reconstructing the signal to emphasize specific features or eliminate noise. The study employed the sym8 wavelet basis for wavelet decomposition and reconstruction. In the denoised signals, a high degree of stability and periodic peaks are distinctly manifested, while amplitude and frequency differences among different types of movements also become noticeably visible. As a result of this study, we are enabled to accurately analyze subtle variations in neck muscle motion signals, such as nodding, shaking the head, neck lateral flexion, and neck circles. Through temporal and frequency domain analysis of denoised motion signals, differentiation among various motion states can be achieved. Overall, this improved analytical approach holds broad application prospects across various types of piezoelectric sensors, such as healthcare monitoring, sports biomechanics.