Model Identification of Distributed Energy Resources Using Sparse Regression and Koopman Theory
Model Identification of Distributed Energy Resources Using Sparse Regression and Koopman Theory
复制标题
DOI：
10.1109/gtd49768.2023.00033
复制
发表时间：
2023-05
期刊：
影响因子：
--
通讯作者：
Javad Khazaei;F. Moazeni
中科院分区：
文献类型：
--
作者：
Javad Khazaei;F. Moazeni
复杂的基于物理的建模过程以及分布式能源内部参数的不确定性和保密性激励了系统识别工具。有了高保真度的测量数据和历史数据，无模型识别法可以简化非线性系统的控制设计，而无需对非线性系统进行繁琐的建模。本文提出了一种基于非线性动力学稀疏辨识（SINDy）的der数据驱动非线性建模框架。此外，为了避免非线性控制设计的复杂性，将利用库普曼理论将识别的非线性动力学提升到线性空间。与现有基于物理的设计严重依赖于了解详细的系统动态或依赖于大量历史数据且不可解释的数据驱动设计相比，所提出的无模型DER识别框架可以通过可用的测量准确捕获DER的动态，并将其提升到线性空间，从而为跟踪问题提供保证性能。通过时域仿真验证了该方法的有效性。
A complex physics-based modeling procedure and the uncertainty and confidentiality of internal parameters of distributed energy resources (DERs) motivate system identification tools. With the availability of high-fidelity measurements and historical data, model-free identification of DERs can facilitate the control design without tedious modeling of these nonlinear systems. This paper develops a framework for data-driven nonlinear modeling of DERs using sparse identification of nonlinear dynamics (SINDy). In addition, to avoid the complexities of nonlinear control designs, the identified nonlinear dynamics will be lifted to a linear space utilizing Koopman theory. Compared with existing physics-based designs that heavily rely on knowing the detailed system dynamics or data-driven designs that relay on large historical data and are not interpretable, the proposed model-free DER identification framework can accurately capture the dynamics of the DERs with available measurements and lift them to a linear space that provides guaranteed performance for tracking problems. Time-domain simulations were carried out to validate the effectiveness of the proposed approach.