From System Identification to Sustainable Offshore Wind Control: Data-Driven and Model-Based Advances
From System Identification to Sustainable Offshore Wind Control: Data-Driven and Model-Based Advances. Seminar of two hours for the University of Science & Technology Beijing - USTB.
Abstract
-
This talk surveys essential directions in system identification and shows how they enable sustainable operation of offshore wind turbines. We revisit core topics, such as nonlinear and hybrid modelling, data-driven and deep-learning approaches, digital twins, and optimal experiment design, together with open challenges in robustness, interpretability, scalability, uncertainty handling, and automation of the identification/validation cycle. These themes set the stage for control and monitoring solutions that combine physical insight with learning, and that integrate seamlessly with predictive control in realistic environments. Building on these principles, we discuss fault detection and isolation as well as fault-tolerant control for modern wind energy systems, with emphasis on offshore deployments. A floating wind benchmark illustrates the discussion, along with farm-level issues such as wake interactions, sensor reliability, model validation, and the role of digital twins for predictive maintenance. The presentation distils practical guidelines for designing identification-aware monitoring and control workflows that are robust, explainable, and ready for industrial adoption.
Keywords
- System identification; data-driven modelling; deep learning for dynamical systems; digital twins; offshore wind turbines; floating wind; fault detection and isolation; fault-tolerant control; predictive maintenance.
Talk's Slides and General Topics
-
Previous Page
Simani Home Page
Dipartimento di Ingegneria