面向模型预测控制的风电机组深度Koopman全局线性建模方法

打开文本图片集

中图分类号：TM614 文献标志码：A

DOI： 10.7652/xjtuxb202602018 文章编号：0253-987X（2026）02-0183-12

Model Predictive Control-Oriented Deep Koopman Global Linear Modeling Method for Wind Turbines

TIAN Runze1，2.3，KOU Peng1，2.3 ，WU Yikun1，2.3， ZHANG Zhihao 1，2，3 ， ZHANG Yuanhang1，23， WANG Ruotao4 ， HAO Shouli4 ， LIANG Deliang1，2，3 （1. School of Electrical Enginering，Xian Jiaotong University，Xi'an 710o49，China;2. Shaanxi Key Laboratoryof Smart Grid，Xi'an Jiaotong University，Xi'an 71O049，China；3. State Key Laboratory of Electrical Insulation andPower Equipment，Xi'anJiaotong University，Xi'an71oo49，China；4.NorthernBranch，Beijing Jingneng International Holding Co.，Ltd.，Hohhot O1000o，China）

Abstract：To address the challenges posed by the complex nonlinear dynamics of wind turbines in model predictive control（MPC）-oriented modeling，as well as the limitations of existing nonlinear and local linear modeling methods in terms of model complexity and accuracy，a global linear modeling method for wind turbines is proposed. Based on Koopman operator theory and deep learning techniques，a statespace mapping neural network is designed. A training strategy incorporating a Frobenius norm-based regularization term is developed to enhance the long-term prediction accuracy of the established model.

Through data-driven training of the proposed network，a high-dimensional global linear dynamic model of the wind turbine is established. Simulation results demonstrate that the prediction errors for rotor speed and pitch angle are 0.869% and 0.026% ，respectively，which are significantly lower than those of the three comparative methods. Compared with the local linear dynamic model，the wind farm MPC strategy based on the established high-dimensional global linear dynamic model reduces the rotor speed tracking error and overshoot by 92.58% and 95.85% ， respectively. The findings provide a theoretical reference for MPC-oriented dynamic modeling of wind turbines.

Keywords： wind turbine；Koopman operator theory；deep learning； global linear modeling；model predictive control

风力发电作为发展最快、最具竞争力的新能源发电技术之一，其装机容量逐年上升[]。（剩余18615字）