Digital Twin for Wind Turbines — Operational Management and AI-driven Predictive Maintenance
In the WindIO research project, CONTACT Software platform technology enable Digital Twin solutions in the wind industry. The project consortium lead by the University of Bremen and CONTACT Software together with seven further industrial partners are working on building up cyber-physical systems, such that smart operational management of the wind turbines is enabled.
Digital twins of two research turbines, a KROGMANN turbine and a SENVION turbine, have been created on a research instance of the platform CONTACT Elements for IoT. The Digital Twin represents the current status of each turbine, like operating and environmental data and supports the prediction of maintenance time windows or the expected energy output. For this, the KROGMANN turbine was equipped with sensors on the three rotor blades that measure parameters such as temperatures and acceleration. Additionally, a lidar was installed to record local weather data at the wind turbine. A vibration box with acceleration sensors at the turbine measures the tower vibration. Via an IoT gateway the data is send to the MQTT broker, which makes it available to the IoT platform. Using those information for condition monitoring, dashboards show the current state of the wind turbines.
To support operational management strategies, dashboards were developed for the digital twin, which provide all users with dashboards for the topics of condition monitoring, weather data and maintenance.
On the one hand, the maintenance dashboard contains supporting widgets, such as a maintenance calendar, in which the upcoming maintenance measures are entered. These are the planned maintenance, but also repair orderings after a malfunction has occurred. The dashboards also show the operating hours, downtimes and the number of events carried out within defined periods. Furthermore, tables with service cases on maintenance and repairs provide information on these activities. Links provide direct access to the objects, which then provide detailed information such as the spare parts required for the repair. A table with a list of all events that have occurred at the wind turbine, such as unplanned downtimes or malfunctions, provides an overview and, by forwarding to the detail page, further insight into the reasons for an event.
The acceleration data from the vibration box is used to predict maintenance windows for e.g. maintenance at the rotor blades [Sander et al.]. For this a machine learning (ML) model has been developed that enables a short time prediction of time windows where maintenance is suitable. The prediction model makes use of the acceleration data from the tower and local weather data. It has been implemented using the SIMWORX low code automation platform from CAIQ that supports a simple access to the operating data from the wind turbine’s digital twin. It is a low code development environment, where the data is accessed by configuration. The ML-code for the prediction model is written in python, which can be integrated into the automation environment easily by configuration.
The model predicts the tower oscillation for up to five minutes. Based on the prediction it is decided whether a maintenance order at the wind turbine is reasonable. This is visualized at the dashboard using the traffic light signal, which is set by an automated alarming of the platform CONTACT Elements for IoT.
A screenshot of a template of the automation integration is shown below. It can be seen that the access to the timeseries data base from the digital twin (asset) can be configured efficiently. Furthermore, the business objects in CONTACT Elements for IoT are automatically created by the elements like events, that are placed by drag and drop into the development environment of the automation.
Please find some more information on the WindIO project on our pages.
References:
[Sander et al.] Sander, A, Haselsteiner, AF, Barat, K, Janssen, M, Oelker, S, Ohlendorf, J, & Thoben, K. “Relative Motion During Single Blade Installation: Measurements From the North Sea.” Proceedings of the ASME 2020 39th International Conference on Ocean, Offshore and Arctic Engineering. Volume 9: Ocean Renewable Energy. Virtual, Online. August 3–7, 2020. V009T09A069. ASME. https://doi.org/10.1115/OMAE2020-18935
