Adaptive Polymorphic Mode Decomposition (APMD)

The matlab codes permit to reproduce some results in the paper: Huang Z and Liu J. Adaptive Polymorphic Mode Decomposition [J]. DSP, 2025.

To extract modes from multi-component multiform complex signal, a framework-like Adaptive Polymorphic Mode Decomposition (APMD) method is put forward in this article. First, Short-Time Fourier Transform (STFT) with optimal window length is applied to obtain the Time-Frequency Representation (TFR) of signal. Then, ridges and bandwidths of each mode are consecutively detected and optimized by iteration. Finally, the signal modes are restored by integration and squeezed in TFR. The idea is simple but novel with combination of Variational Mode Decomposition (VMD)-like methods and Synchro Squeezing Transform (SST)-like methods, which is non-parameterized and fully adaptive. Results of decomposing some typical signal verify the effectiveness and robustness in analyzing complex polymorphic signals, being more suitable than traditional methods for decomposing signals mixed with both time-dominant and frequency-dominant components.