Wind Turbine Monitoring Using Optical Motion Magnification: Challenges and Opportunities

Abstract

An inexpensive, quick, and robust structural health monitoring technique to assess the integrity of wind turbine foundations and components remains elusive. Traditional techniques for wind turbine condition monitoring (CM) are costly, invasive, and cause downtime in the asset. As turbines age, wear-induced damage may accumulate in the structural and mechanical components. The ability to effectively monitor wind turbine foundations and blades using optical methods is explored in this work. Specifically, optical motion magnification (OMM) is investigated for this application. OMM is a computer vision-based monitoring technique that can extract and amplify unperceivable displacements in a video of a structure of interest. These magnified videos effectively reveal the areas in a system with the largest displacements. OMM is a remarkable visualization tool, giving technicians the ability to quickly identify areas with problematic vibrations in an easily understood manner. OMM also generates quantitative displacement time histories from which the amplitude and frequency spectra of the structure can be extracted. This work describes two experimental studies completed to determine the feasibility of OMM when this approach is used to inspect wind turbine foundations and blades. Firstly, an experiment exploring the use of OMM for modal analysis of turbine blades was performed, comparing the frequency spectra calculated using both optical and traditional contact-based measurements. OMM proved capable of identifying resonant frequencies with an error less than 2.5% when compared to conventional measurements. Next, proof-of-concept turbine foundation tests were completed to ascertain the viability of the method to distinguish between questionable and healthy foundations. The results of this research show how OMM can potentially eliminate several issues inherent in standard testing and monitoring procedures, namely reducing the overall cost of the sensor systems and time spent collecting data. If further developed, OMM can open the doors to many more innovative solutions, applications, and optimizations of current CM practices for large-scale engineering systems.