arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2409.11586eess.SYcs.SY
收藏

基于不完全测量的分布式Koopman学习

Distributed Koopman Learning with Incomplete Measurements

School of Aeronautics and Astronautics at Purdue University(普渡大学航空航天学院)

Wenjian Hao, Lili Wang, Ayush Rai, Shaoshuai Mou

AI总结:

针对观测能力有限的智能体网络协同估计非线性时不变系统动力学的问题,提出融合Koopman算子理论、深度神经网络和共识协调的分布式深度Koopman学习算法,仅靠局部测量和邻居交互即可准确重构全局动力学,性能接近全状态集中式方法。

摘要:

Koopman算子理论已成为系统辨识的有力工具,尤其适用于逼近非线性时不变系统(NTIS)。本文研究由观测能力有限的智能体组成的网络协同估计NTIS动力学的问题。通过融合Koopman算子理论、深度神经网络和基于共识的协调机制,提出了一种分布式深度Koopman学习算法。在该框架中,每个智能体利用自身的局部测量值和与邻居交换的提升状态来逼近系统动力学。这种协同方案即使在单个智能体不具备全状态信息的情况下,也能准确重构全局动力学。在OpenAI Gym的Lunar Lander环境上的仿真结果表明,该方法的性能与可访问全状态的集中式深度Koopman学习相当。

英文摘要:

Koopman operator theory has emerged as a powerful tool for system identification, particularly for approximating nonlinear time-invariant systems (NTIS). This paper considers a network of agents with limited observation capabilities that collaboratively estimate the dynamics of an NTIS. A distributed deep Koopman learning algorithm is developed by integrating Koopman operator theory, deep neural networks, and consensus-based coordination. In the proposed framework, each agent approximates the system dynamics using its partial measurements and lifted states exchanged with its neighbors. This cooperative scheme enables accurate reconstruction of the global dynamics despite the absence of full-state information at individual agents. Simulation results on the Lunar Lander environment from OpenAI Gym demonstrate that the proposed method achieves performance comparable to the centralized deep Koopman learning with full-state access.

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。

↑