Snake Optimization-Based Variable-Step Multiscale Single Threshold Slope Entropy for Complexity Analysis of Signals

Slope entropy (SloEn) is an effective complexity analysis measure of signals that has been applied to many areas in recent years. Whereas SloEn can only reflect the complexity information of signal at a single scale and SloEn is sensitive to its threshold value, the choice of the threshold value will affect the SloEn value. In this article, the variable-step multiscale single threshold SloEn (VSM-StSloEn) is proposed, not only reflecting the complexity information hidden in different time scales but also compensating for the shortcomings of traditional multiscale processing, which is based on the recently proposed single threshold SloEn (StSloEn) that simplifies the calculation process of SloEn and combines variable-step multiscale processing. Furthermore, to overcome the threshold selection problem of VSM-StSloEn, this article proposes snake optimization-based VSM-StSloEn (SO-VSM-StSloEn) by introducing snake optimizer (SO). Finally, we have verified the performance of SO-VSM-StSloEn on simulated and real-world experiments, and the results demonstrate that SO-VSM-StSloEn is insensitive to signal length, independent on the threshold and has stronger distinguishing capability than other similar improved algorithms based on SloEn; furthermore, SO-VSM-StSloEn is superior to other commonly used entropies in classifying different categories of real-world signals.