An Improved Envelope Fitting Algorithm for Frequency Spectrum Analysis
作者
Lingyun Zhang,Hui Tan
标识
DOI:10.1109/aces-china56081.2022.10065376
摘要
In order to solve the problem of undershoot and non-monotonicity in the envelope fitting applied to frequency spectrum analysis, this paper proposed an improved envelope fitting algorithm based on linear interpolation. The improved algorithm performs envelope fitting on the non-stationary region of the original signal by combining the method of finding extreme values and linear interpolation, and rejects the non-monotonic extreme points to ensure the monotonicity of the envelope. For the stationary signal regions, the envelope can tightly wrap the original signal and retain the characteristics of the original signal. The simulation results shows that the method proposed in this paper can effectively fit the envelope of the original signal without undershoot regions.