Wind Turbines with Optimized Productivity through Fleet Monitoring without Additional Sensor Technology
Abstract
The paper is a brief presentation of a recently started research project (WEAproduktiv) on optimizing the electricity production of wind turbines (WT) by means of population or fleet monitoring. The project is not mainly about Structural Health Monitoring (SHM), but about finding the causes of suboptimal power production. Nevertheless, damages, manufacturing defects or inaccuracies as well as suboptimal control, etc. are some of the reasons, which can lead to a loss of power production. We assume that e.g. causes like damages, manufacturing defects or a poorly functioning of the control system of a WT are reflected to some extent in the vibration behavior of the plant. As shown below, there exist many causes for production losses. They can be most easily detected by means of population monitoring. The objective is to identify “suspect” plants in wind parks and the causes of their power production losses using only Supervisory Control and Data Acquisition (SCADA). But first, the relationships between SCADA, vibration data and power productivity must be better understood. To achieve this, a numerical (digital twin) and a data driven model (physical twin) of a socalled fleet leader are included in our calculations. In a second step, the digital twin and the data driven model will be extended to understand and to model the population of wind power plants in different wind parks. The decision whether the intended population monitoring is possible with SCADA data alone and therefore the vibration data can be omitted, will be made towards the end of the research project. This depends on the meaningfulness of the generated models.
DOI
10.12783/shm2023/36772
10.12783/shm2023/36772
Full Text:
PDFRefbacks
- There are currently no refbacks.