Posters
Come meet the poster presenters to ask them questions and discuss their work
We would like to invite you to come and see the posters at our upcoming conference. The posters will showcase a diverse range of research topics, and will give delegates an opportunity to engage with the authors and learn more about their work. Whether you are a seasoned researcher or simply curious about the latest developments in your field, we believe that the posters will offer something of interest to everyone. So please join us at the conference and take advantage of this opportunity to learn and engage with your peers in the academic community. We look forward to seeing you there!
PO272: Verification of a Machine Learning approach to calibrate local wind speed measurements using neighbouring turbines.
Philip Bradstock, Head of Analytics, Bitbloom
Abstract
Accurate and reliable measurement of local wind speed is critical in wind energy performance analysis. However, time-variant bias in nacelle anemometry is a well-known and widespread challenge in the field leading to inaccuracy and uncertainty. This makes observation of apparent performance changes through time variant nacelle anemometer power curves (NAPC's) prone to false positives. Here, we present a novel machine learning based method that allows us to recalibrate local wind speed measurements of a turbine using more reliably measured operational SCADA data from neighbouring turbines. Additionally, following removal of clear outliers/stuck/invalid data, the method is not dependent on any identification of turbine operational modes, nor measurement of any wind speeds for inference. The trained model produces predictions of wind speed which is used to recalibrate the nacelle anemometer wind speed signal. This recalibrated signal becomes a more consistent estimate of wind speed and is used to build calibrated NAPC's. This provides fast and robust identification of variations in performance over time with fewer false positives. Here, we demonstrate the application of the method on a range of sites across Europe, to act as a verification of the reliability of the approach. We show that the method is consistent across varying wind farm sizes and layouts, and that the method is generally robust to terrain and waking effects.
