A multisynchrosqueezing-based high-resolution time-frequency analysis tool for the analysis of non-stationary signals

• A concentrated time-frequency analysis method is proposed in the study. • The proposed method allows for perfect signal reconstruction. • Experimental examples are employed to validate the proposed method. In this paper, a high-resolution time-frequency (TF) analysis method is presented for the analysis of strongly non-stationary signals. TF representations generated by conventional methods are usually too blurry to provide precise features for such signals. A recently proposed method, called multisynchrosqueezing transform (MSST), overcomes most of the problems that exist in conventional methods, which seems to be a promising tool. However, the MSST still has a major problem, i.e., non-reassigned point problem, which may lead to the blurry energy problem for some special TF points. This paper mainly focuses on resolving this problem. This study finds that such a problem in the MSST is caused by the rounding operation in the discrete procedure of the reassigned step. An effective method is then employed to address this problem using a simple strategy. Additionally, discrete implementation is provided in the study. The numerical analysis shows that our proposed method can effectively improve the energy concentration comparable to the MSST. Comparisons with other advanced methods also show that the proposed method offers better performance in addressing strongly non-stationary signals and noise-added signals. In the experimental signal analysis, we carry out three experiments to validate the effectiveness of the proposed method in the analysis of real-world signals.