Spring Test for Automatic Measuring of Signal Complexity

Jie Cheng,Haiyong Huang,Wenshi Li

DOI: https://doi.org/10.21203/rs.3.rs-1552057/v1

2022-01-01

Abstract:Abstract In order to automatically measure the complexity of time series, we invented 3S plot, and the embedded spring is a one-dimensional manifold. Three automatic statistical indexes are defined by self-similarity measure, extreme value search acceleration and neighborhood rotation phase, respectively. The above three new metrics and another two comparison standards, Lyapunov exponent and spectral entropy complexity, can pass the benchmark system test of unified equation. Under the guidance of visualization of bifurcation diagram, three probability measuring methods with relatively weak independent recognition abilities and one phase measuring method with relatively sensitive detection ability can easily identify the order and chaos states following Chua's equation. Finally, five main measuring tools are used to successfully characterize the new 555 timer-based chaotic signals.