arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2409.07642cs.LGcs.SYeess.SY
收藏

使用System Identification Toolbox(TM)的动态系统深度学习

Deep Learning of Dynamic Systems using System Identification Toolbox(TM)

MathWorks(MathWorks公司);Jordan University of Science and Technology(约旦科学技术大学)

Tianyu Dai, Khaled Aljanaideh, Rong Chen, Rajiv Singh, Alec Stothert, Lennart Ljung

AI总结:

本文介绍近三年MATLAB的System Identification Toolbox(TM)在动态建模上的进展,包括整合深度学习技术、推出可扩展自编码的神经状态空间模型,及多项机器学习集成、状态估计与训练数据格式相关增强功能。

摘要:

过去三年间,MATLAB(R)版本迭代中,System Identification Toolbox(TM)提供的动态建模能力持续增强。其重点在于整合深度学习架构与训练技术,推动深度神经网络作为非线性模型的构建模块使用。该工具箱提供神经状态空间模型,可通过自编码特性扩展,尤其适用于大型系统的降阶建模。此外,工具箱还有多项其他增强功能:深化与前沿机器学习技术的集成、利用自动微分特性进行状态估计,以及支持直接使用原始数值矩阵和时间表训练模型。

英文摘要:

MATLAB(R) releases over the last 3 years have witnessed a continuing growth in the dynamic modeling capabilities offered by the System Identification Toolbox(TM). The emphasis has been on integrating deep learning architectures and training techniques that facilitate the use of deep neural networks as building blocks of nonlinear models. The toolbox offers neural state-space models which can be extended with auto-encoding features that are particularly suited for reduced-order modeling of large systems. The toolbox contains several other enhancements that deepen its integration with the state-of-art machine learning techniques, leverage auto-differentiation features for state estimation, and enable a direct use of raw numeric matrices and timetables for training models.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑