Adaptive Time-Frequency-Supported Chirp Component Decomposition.
Gang Shi,Chengjin Qin,Zhinan Zhang,Jianfeng Tao,Chengliang Liu
DOI: https://doi.org/10.1109/tim.2023.3323958
IF: 5.6
2023-01-01
IEEE Transactions on Instrumentation and Measurement
Abstract:There are a lot of complex amplitude modulated-frequency modulated (AM-FM) signals with strong noise in the real world. The decomposition of complex signals can effectively reveal the variation laws of each subsignal. However, current methods cannot effectively decompose complex signal with strong noise. To solve this problem, we propose a novel adaptive time-frequency-supported chirp component decomposition (ATCCD) in this article. ATCCD establishes a new adaptive signal decomposition approach by constraining the smoothness of demodulated signal and the subsignal decomposition error. Then, ATCCD establishes a new instantaneous frequency (IF) update optimization model by integrating time-frequency transform (TFT) iterative optimization and IF arctangent demodulation optimization. The new IF update model enhances the level of IF iterative optimization with the support of signal's time-frequency feature information, which can effectively filter out the interference of strong noise and accurately estimate IF. ATCCD bridges the gap between signal decomposition and TFT, makes them promote each other's performance, and effectively solves the mode aliasing problem. In this way, ATCCD can accurately estimate the IFs and reconstruct subsignals for complex signal with strong noise. Furthermore, the decomposition results verify that ATCCD can more accurately decompose complex AM-FM signals with strong noise containing crossing IFs, close Ifs, or fast changing IFs than current signal decomposition and TFT methods. Simultaneously, ATCCD can more precisely decompose complex experimental signals with strong noise and obtain higher quality time-frequency representations (TFRs).