An Improved Method to Enhance SNR of Non-Stationary Signal Based on Nonuniform Windows Adaptive Modulation for Measurement.
Lidong Huang,Xinglin Sun,Tiantian Huang,Zhijuan Zhu,Kaichen Song
DOI: https://doi.org/10.1109/tim.2023.3279875
IF: 5.6
2023-01-01
IEEE Transactions on Instrumentation and Measurement
Abstract:It is very important for non-stationary signal processing, signal measurement, and signal enhancement to improve signal-to-noise ratio (SNR) and suppress noise. For complex non-stationary signals measurement, there is still a lack of effective methods to improve the SNR without damaging original signal. Therefore, it is vital to develop signal processing method to improve SNR of non-stationary signal in many actual industrial measurements. This article proposes a new method to enhance the SNR for non-stationary signal processing. First, a self-adapting mathematic model is presented to describe the short-time impulse with board spectrum (SIBS), judge the non-stationary part of the signal, and determine the length of time window based on itself. Then, for the non-stationary signals in this short time window, the improved self-adaptive modulation is used to remove the non-stationary components and retain the stationary components. Third, the stationary component of the short-time window and the normal part of the original signal are connected via cubic spline interpolation algorithm. Finally, this article analyzes the range of SNR improvement and measurement accuracy of the new method. Based on the results of experiment, the proposed method can effectively improve the SNR of non-stationary signals and suppress noise, which is of great significance for measurement.