Wind Farm Cooperative Yaw Predictive Control Against Model Uncertainty via Continuous Learning From SCADA Data

View at publisher Summarise with VIRAT
Authors: Yu Wang, Shanbi Wei, F. Zhang, Wei Yang, Yi Chai
Cited by: 0
References: 43

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Cooperative yaw control is a critical solution to reduce wake effect and improve power generation of wind farms, and its implementation based on analytical wake model is crucial to realizing its engineering practice. However, the inherent model uncertainty raises challenges. To this end, this paper proposes an application-oriented framework that allows SCADA data to engage cooperative yaw control, which can improve model adaptability and ensure power gain. First, considering the operation principle of industrial yaw system, a novel cooperative yaw predictive control with yaw rate as control action is introduced. Importantly, a continuous learning strategy against model uncertainty is proposed, which is composed of three mechanisms: outlier online detection using model prior knowledge, working scenario generation by distance dependent Chinese restaurant processes, and parameter learning with Bayesian inference. Finally, the result derived from Penmanshiel wind farm demonstrates that the proposed strategy can assimilate SCADA data well with a promising potential for engineering application of model-based cooperative yaw control.

Fields of Study & Subjects

Wind Turbine Control Systems Wind Energy Research and Development Advanced Control Systems Optimization

Publication details

Journal / source
IEEE Transactions on Sustainable Energy
Volume
17
Issue
4
Pages
3496–3507
Publication date
26 March 2026
Language
EN