Online Monitoring of Wind Power Plants using Digital Twin Models

In recent years, there have been significant developments of offshore wind technology and industry, with bottom-fixed wind turbines fully commercialized and floating wind turbines entering the market. Reducing Operational Expenditure (OPEX) for offshore wind turbines by improving the wind turbine availability based on predictive maintenance of the turbine critical components can contribute substantially to the reduction of unexpected maintenance and costs and the development of more sustainable offshore wind energy in future. For this purpose, digital twin models are an enabler. Challenges of digital twin models for predictive or condition-based maintenance mainly arise from:

• Finding the minimum model to capture the various dynamics and relevant failure mechanisms for a wide range of wind turbine, wind farm and the support grid components.

• Obtaining the data processing algorithms and continuously processing architectures which are able to deal with the real-time aspects of digital twin models, and optimizing heterogeneous data exchange between different models in the digital twin framework.

• Developing solutions such as knowledge graph databases and domain ontologies for ensuring efficient data handling for continuous, automated and semi-automated updating of digital representations of the connected assets.

• Identifying the sources of model and measurement uncertainties in the different steps of digital twin model realization and mitigating their influences.

It is our pleasure to invite researchers working on all the different aspects of the realization of digital twin models for predictive maintenance of on-and offshore wind turbines and power plants to send their contributions to this research topic in the Wind Energy section of the journal of Frontiers in Energy Research. This Research Topic welcomes the original research and review articles which contribute to the key technical challenges of digital twin models for intelligent monitoring of the structural, rotating, electrical and electronic systems and components of wind turbines (onshore or offshore, bottom-fixed or floating) and power plants, which includes (but is not limited to) using the computationally efficient models, model updating algorithms that can be executed in real-time, stochastic modelling and statistical techniques to handle uncertainties. This Research Topic covers the three layers of digital twin models realization, namely data acquisition and transmission layer (sensors and data collection); storing and processing data through the platform layer (dynamic and degradation models); and decision support through the application layer (maintenance planning and optimization).

This Research Topic invites various forms of contribution to digital twin modelling of wind power plants for predictive maintenance. The areas of interest include (but not limited to):

• Wind turbine and farm analytical and numerical modelling

• Inverse engineering based on response measurements

• Distributed and decentralized computing for real-time estimation of model parameters

• Surrogate modelling of turbine responses

• Signal processing for signal denoising and feature selection

• Statistical approaches and stochastic models for modelling and mitigating uncertainties

• Physics- and machine learning-based degradation models

• Structural health monitoring of wind turbines

• Health monitoring of power electronics, cables and other grid elements

• Power train system condition monitoring

• SCADA-based condition monitoring

• Sensors and data acquisition

• Optimal maintenance management

• Tuning wind farm and turbine controllers by using digital twin models

• Development of digital twins to optimize the exploitation of wind farms and farm clusters

Keywords: digital twin, wind turbines, dynamic modeling, load observer, remaining useful life, model estimation, predictive maintenance, condition monitoring, structural health monitoring, cable health monitoring, grid health monitoring, power electronics health monitoring, power train, uncertainty modeling, generation forecasting, artificial intelligence, distributed

Important note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.