Improving turbine park performance through predictive maintenance.
Even well-running turbines experience problems once in a while. The question therefore is how turbine owners can predict turbine errors and avoid consequential production loss and costly unplanned service visits.
PRO Engine is a real-time data-handling engine for prediction, reliability and optimisation algorithms. The platform enables detection of problems before they result in wind turbine stop so predictive maintenance can be performed which in turn increases availability of the wind turbine.
The platform comes in different versions; either fully equipped with turbine-specific algorithms ready for predictive maintenance or for the customers to add algorithms themselves.
PRO Engine includes Python and Matlab™/Octave algorithm development support. The platform has a dynamic engine where additional algorithms can be directly deployed without any deep software knowledge from the customer’s side.
The platform provides easy overview of the individual wind turbine components’ health status by use of the Pro Engine monitor. From the turbine health monitor, the customer can get information about possible root causes and which spare parts and tools to bring.
Finally, PRO Engine contains detailed information of the individual wind turbine components for the customer to observe the measurement and model data that underlie the health calculations.
We use our high expertise within wind to develop algorithms and we apply machine learning methods to calculate a health value between 100% and 0%.
A value of 100% indicates a perfectly healthy system as opposed to 0%, which indicates a system failure.
When the value reaches 50% it is indicated on the PRO Engine surveillance as yellow and when 20% is reached it gets red
This approach allows monitoring of fairly complex systems, and predictive maintenance can be planned just by reacting according to the colours.
Read more about the benefits and customer projects in our PRO Engine brochure:
