Scholarlink

<p>This paper introduces an online tensor-based dynamic mode decomposition (DMD) method for high-dimensional time-varying systems. By leveraging the tensor train (TT) decomposition, our approach efficiently handles the scalability challenges associated with traditional DMD methods. The proposed framework not only reduces computational complexity but also enables adaptive analysis of systems with changing dynamics. We demonstrate its application in high-dimensional optimal control, where it effectively identifies dominant modes and dynamics, providing ideas for designing more effective control strategies. Numerical experiments validate the effectiveness and highlight its potential for addressing complex control tasks in various fields.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Online Tensor-Based Dynamic Mode Decomposition for Time-Varying System

  • 期刊论文

  • 发表日期：

    2026年5月19日

查看全文
  • Wanli Ma,
  • Chengdong Liu,
  • Yimin Wei

摘要

This paper introduces an online tensor-based dynamic mode decomposition (DMD) method for high-dimensional time-varying systems. By leveraging the tensor train (TT) decomposition, our approach efficiently handles the scalability challenges associated with traditional DMD methods. The proposed framework not only reduces computational complexity but also enables adaptive analysis of systems with changing dynamics. We demonstrate its application in high-dimensional optimal control, where it effectively identifies dominant modes and dynamics, providing ideas for designing more effective control strategies. Numerical experiments validate the effectiveness and highlight its potential for addressing complex control tasks in various fields.