TY - JOUR
T1 - High-fidelity displacement reconstruction method for offshore wind turbines under extreme sea conditions and high noise
AU - Wei, Ziwen
AU - Hao, Zhigang
AU - Lv, Shiyan
AU - Chen, Ding
AU - Zhang, Meng
AU - Zhang, Xiaohui
AU - Xie, Zhongliang
N1 - Publisher Copyright:
Elsevier Ltd
PY - 2026/1/1
Y1 - 2026/1/1
N2 - Conventional methods, such as modal expansion, orthogonal basis superposition, and the inverse finite element method, struggle with poor noise robustness when reconstructing wind turbine tower responses under non-stationary wind loads and high measurement noise in complex sea conditions. A reconstruction methodology integrating dynamic-static hybrid basis functions with an augmented Kalman filter (DS-HB-AKF) is developed herein as a resolution to these identified limitations. The DS-HB-AKF method inherently improves reconstruction accuracy under high measurement noise through a three-step framework. First, the method constructs a hybrid basis using fourth-order modal and second-order quasi-static basis functions to establish the observation matrix and initial system model. Second, a sequential greedy algorithm minimizes the observation matrix's condition number, directly optimizing sensor placement. Finally, the augmented Kalman filter applies optimal recursive estimation to synchronously track and precisely decouple unknown quasi-static coefficients and modal coordinates from noisy observation sequences. Validation across 27 finite element simulation cases demonstrated the method's high fidelity. The relative root-mean-square error (RRMSE) of the full-field displacement reconstruction remained strictly below 18.5%, even under the most adverse combination of a 10 dB signal-to-noise ratio and a 24 m/s wind speed. The proposed method offers a potential underlying algorithm for the full-field state awareness of large offshore wind structures using limited physical measurement points.
AB - Conventional methods, such as modal expansion, orthogonal basis superposition, and the inverse finite element method, struggle with poor noise robustness when reconstructing wind turbine tower responses under non-stationary wind loads and high measurement noise in complex sea conditions. A reconstruction methodology integrating dynamic-static hybrid basis functions with an augmented Kalman filter (DS-HB-AKF) is developed herein as a resolution to these identified limitations. The DS-HB-AKF method inherently improves reconstruction accuracy under high measurement noise through a three-step framework. First, the method constructs a hybrid basis using fourth-order modal and second-order quasi-static basis functions to establish the observation matrix and initial system model. Second, a sequential greedy algorithm minimizes the observation matrix's condition number, directly optimizing sensor placement. Finally, the augmented Kalman filter applies optimal recursive estimation to synchronously track and precisely decouple unknown quasi-static coefficients and modal coordinates from noisy observation sequences. Validation across 27 finite element simulation cases demonstrated the method's high fidelity. The relative root-mean-square error (RRMSE) of the full-field displacement reconstruction remained strictly below 18.5%, even under the most adverse combination of a 10 dB signal-to-noise ratio and a 24 m/s wind speed. The proposed method offers a potential underlying algorithm for the full-field state awareness of large offshore wind structures using limited physical measurement points.
KW - Augmented Kalman filter
KW - Displacement reconstruction
KW - Hybrid basis functions
KW - Structural health monitoring
KW - Wind turbine tower
UR - https://www.scopus.com/pages/publications/105051347777
U2 - 10.1016/j.oceaneng.2026.128360
DO - 10.1016/j.oceaneng.2026.128360
M3 - 文章
AN - SCOPUS:105051347777
SN - 0029-8018
VL - 368
JO - Ocean Engineering
JF - Ocean Engineering
M1 - 128360
ER -