A Directional Time-Frequency Distribution Instantaneous Frequency Estimation Algorithm Based on Improved Dynamic Programming Algorithm

When the potential time-frequency distribution is affected by components of the signal intersecting in the time-frequency space, algorithms for estimating instantaneous frequency based on time-frequency analysis often fall short of the desired performance. This article uses a general linear chirplet transform to obtain a more concentrated time-frequency representation of energy, and introduces a modified version of the dynamic programming algorithm that leverages both the direction and amplitude of signal elements to determine the instantaneous frequency of overlapping components, effectively addressing the aforementioned issue.