To address issues such as low signal-to-noise ratio and complex noise types in static measurements of force sensors, this paper proposes a noise reduction method based on
the fusion of two-layer wavelet decomposition and LMS adaptive filtering. This method first employs median filtering to eliminate isolated impulse noise as preprocessing. Subsequently, it utilizes Daubechies4 (db4) and Symlets8 (sym8) wavelet bases for two-stage cascaded wavelet decomposition and thresholding, effectively suppressing multi-frequency noise. Following this, an LMS adaptive filter intelligently suppresses residual noise. Experimental results demonstrate that this method outperforms
traditional single-wavelet denoising approaches in terms of signal-to-noise ratio (SNR) and root mean square error (RMSE). It achieves significant noise reduction while preserving signal morphological features, indicating its feasibility for processing force sensing signals.
发表评论