多元统计
熵(时间箭头)
非线性系统
传递熵
因果关系(物理学)
混乱的
计算机科学
统计物理学
算法
数学
人工智能
最大熵原理
物理
机器学习
热力学
量子力学
作者
Xingran Li,Chunling Fan,Jiangfan Qin,Rui Yang
标识
DOI:10.1515/zna-2023-0115
摘要
Abstract This paper presents a refined composite multivariate multiscale complexity-entropy causality plane (RCMMCECP) to explore the dynamics features of gas–liquid two-phase flow. Firstly, we employ a series of typical nonlinear time series to confirm the effectiveness of the RCMMCECP, including seven chaotic systems, two random processes, and one periodic process. The comparison results of the proposed method and conventional multivariate multiscale complexity-entropy causality plane (MMCECP) confirm the stability performance of the proposed RCMMCECP. Above all, the RCMMCECP enhances the reliability of the statistical complexity measure over large time scales and exhibits good continuity and noise-resistant ability in multiscale analysis. Then, we employ the RCMMCECP to analyze the upstream and downstream conductance signals. The experimental results demonstrate that the RCMMCECP can characterize the change of complexity and structural stability in the gas-liquid two-phase flow evolution process, effectively revealing its dynamics features.
科研通智能强力驱动
Strongly Powered by AbleSci AI