Diagnostic Digital Twin for Anomaly Detection in Floating Offshore Wind Energy

TL;DR

The paper tackles the challenge of real-time health monitoring for remote floating offshore wind assets by developing a diagnostic digital twin that uses unsupervised normal-operation models (NOMs) to detect anomalies and diagnose faults. It compares dense neural networks (DNN) and long short-term memory networks (LSTM) as NOMs, employs SHAP explanations for root-cause attribution, and integrates results into a Unity-based VR interface with SMS alerts. On data from an operational turbine, the DNN NOM detected a temperature anomaly hours before a prolonged downtime, with SHAP identifying the generator rotor and stator as the fault source; the LSTM NOM demonstrated less clear anomaly signals. The approach is generalizable to other offshore assets and aims to increase asset lifetime, efficiency, and sustainability, though broader validation across multiple turbines and real-time industrial deployment are needed to build trust and enable scale.

Abstract

The demand for condition-based and predictive maintenance is rising across industries, especially for remote, high-value, and high-risk assets. In this article, the diagnostic digital twin concept is introduced, discussed, and implemented for a floating offshore turbine. A diagnostic digital twin is a virtual representation of an asset that combines real-time data and models to monitor damage, detect anomalies, and diagnose failures, thereby enabling condition-based and predictive maintenance. By applying diagnostic digital twins to offshore assets, unexpected failures can be alleviated, but the implementation can prove challenging. Here, a diagnostic digital twin is implemented for an operational floating offshore wind turbine. The asset is monitored through measurements. Unsupervised learning methods are employed to build a normal operation model, detect anomalies, and provide a fault diagnosis. Warnings and diagnoses are sent through text messages, and a more detailed diagnosis can be accessed in a virtual reality interface. The diagnostic digital twin successfully detected an anomaly with high confidence hours before a failure occurred. The paper concludes by discussing diagnostic digital twins in the broader context of offshore engineering. The presented approach can be generalized to other offshore assets to improve maintenance and increase the lifetime, efficiency, and sustainability of offshore assets.

Figures (12)

• Figure 1: Digital twin framework and capability level scale Stadtmann2023doa.
• Figure 2: Dense neural network structure
• Figure 3: Long short-term memory neural network structure
• Figure 4: Anomaly detection algorithm using a normal operation model and taking into account missing sensor values, turbine downtime, reheating process, and extended reheating. For more details see Section \ref{['ssec:m_anom']}.
• Figure 5: Digital twin virtual reality interface with components colored according to the measured temperature.