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!
PO006: AI-based models for the real-time detection of power curve performance deviations
Carlos Niederbacher, Product Manager, Fluence
Abstract
Asset management generally requires large numbers of turbines to be monitored in parallel, and digital tools to aid the early detection of performance issues are increasingly necessary for an efficient monitoring process. Real-time automatic detection of power curve performance deviations of wind turbines helps asset managers and owners to better prioritize, reduce reaction time and limit losses. We present a comprehensive solution that uses AI-based models applied to SCADA data to support and improve power curve performance deviation monitoring. The solution includes models running in real time that identify eventual wind turbine performance deviations. Users are notified by an alarming system in the shortest time possible, allowing them to inspect and react. The expected turbine performance is the output of the AI-based models. Such models are turbine specific and trained on a subset of its historical SCADA data. The training set must be representative of all possible operational conditions and concurrently it must be cleared of any periods of time when the turbine underperformed. Advanced filtering logic based statistical models, in addition to the information provided by the SCADA (I.e. turbine status or mode) are used at this scope. The alarm system logic determines when detected power curve performance deviation shall trigger an alarm to the user. The solution has been delivered to clients and field data results showed that the designed alarm system successfully detects relevant power deviations that cause performance issues, I.e. caused by icing, curtailments, yaw misalignment, pitch control related. Furthermore, it proved to be a successful tool in reducing the reaction time and identifying issues that could not be detected in short term or were undetectable before the solution was activated. Model is OEM agnostic thus highly scalable, and it can be quickly deployed on large fleets of turbines composed of different manufacturers.
