TY  - EJOU
AU  - Luo, Chao 
AU  - Guo, Ziteng 
AU  - Zeng, Guanming 
AU  - Zhu, Yihua 
AU  - Zhang, Chen 

TI  - A Data-Driven Feature Extraction Method for Grid Integration Analysis and Reduced-Order Modeling of Floating Offshore Wind Turbines
T2  - Energy Engineering

PY  - 
VL  - 
IS  - 
SN  - 1546-0118

AB  - Effective reduced-order models (ROMs) are in urgent demand for analyzing the dynamic features of floating offshore wind turbines (FOWTs), such as ROMs for control and grid-integration analysis of FOWTs-based wind parks. However, due to the complexity of the coupled dynamics of FOWTs, the identification and the choice of dominant features are crucial for ROMs. To achieve rational guidelines for developing such ROMs, this paper proposes the quantitative and data-driven approaches to extract the dominant degrees of freedom (DOFs) of FOWTs. First, the load characteristics of FOWT and the spatiotemporal features of DOFs under external excitations are analyzed. Then, the vertical and lateral DOFs of FOWT are divided based on these features, in which vertical DOFs belong to nonstationary slow dynamics, and lateral DOFs belong to multiscale fast dynamics. Based on this, data-driven approaches, including the Spearman correlation analysis and dynamic mode decomposition (DMD), are utilized to identify and extract the dominant mechanical components and associated DOFs under nonstationary slow dynamics and multiscale fast dynamics, respectively. The presented method is validated using simulation data from the OpenFAST-MATLAB/Simulink co-simulation platform. The results show that the proposed data-driven approaches can effectively extract the dominant DOFs of FOWT relevant to dynamic grid-integration analysis.
KW  - Floating offshore wind turbine; grid-tied power generation; Spearman correlation analysis; dynamic mode decomposition; degrees of freedom

DO  - 10.32604/ee.2026.081600