Research on Improved Regularization Method Based on Signal Waveform Measurement

Abstract

Aiming at the ill-posed problem of deconvolution in signal waveform measurement, an improved method based on Tikhonov regularization and TSVD regularization is proposed, combining Tikhonov regularization and TSVD regularization, the singular value is divided into large singular value and small singular value by setting threshold, and the larger singular value is retained, at the same time, Tikhonov regularization is used to modify the smaller singular value. In this method, the effective information of signal waveform is preserved as much as possible, and the effect of noise on deconvolution is reduced, so that more accurate reconstructed signal waveform can be obtained. The simulation research is carried out in detail, and the signal reconstruction is realized, and the more accurate signal waveform is obtained. The results show that the improved Tikhonov regularization method can effectively solve the ill-posed problem of deconvolution in signal waveform measurement.