Search

Browse

• Communities
• Authors
• Year
• Titles
• Subjects

My Space

• Login

Predictive maintenance and fault detection on onshore windfarm using digital twins

Kapenis Antonios

Full record

---

Summary

This thesis places its focus on the development of a digital twin that faithfully embodies a physical wind farm located in Greece. The principal objective is to establish a virtual counterpart that emulates the real-world characteristics and dynamics of the wind farm. In order to accomplish this, the thesis presents algorithms that are specifically devised to facilitate three vital functionalities: power output prediction, predictive maintenance and fault detection. These algorithms are an integral part of the digital twin's operation, enabling it to forecast potential issues and identify existing problems in the wind turbines. An important characteristic of the digital twin devised in this thesis is its capability to regulate the operations of the wind turbines, per demand. This entails monitoring their performance and, crucially, taking appropriate measures in the event of a malfunction. When the system recognizes a malfunction, it possesses the capability to either temporarily or permanently suspend the operation of the affected turbines until the issue is completely resolved. This approach ensures that any problems are promptly addressed, minimizing downtime and potential harm. A crucial element of the wind farm digital twin centers around the incorporation of real-time data acquired from the wind farm. This data is essential in order to execute the algorithms, as it provides the vital input for the digital twin to successfully perform its functions of predictive maintenance and fault detection. Through the utilization of genuine operational data, the digital twin can generate more accurate predictions and diagnoses, ultimately resulting in a more effective and dependable management of the wind farm.

Available Files

Services