Damage Identification of Offshore Wind Turbine Blades Considering Material Aging Based on NFCRR and Stacking Model

Accurate and reliable detection of damage to offshore wind turbine blades (OWTBs) can enhance the reliability and safety of wind power generation systems. This article proposes a novel method to accurately identify damage in OWTBs by using the ratio of any two natural frequency change rates (NFCRRs) and stacking machine learning techniques. Compared with previous damage identification methods, this approach addresses the challenge of material aging, which is crucial for improving the long-term reliability of damage identification systems. By introducing the NFCRR as a novel damage indicator, the method minimizes the effects of material aging and improves the long-term effectiveness of damage identification. Furthermore, stacking machine learning is adopted to enhance the accuracy of damage identification. The effectiveness of this method is verified by taking crack damage identification on composite cantilever beams equivalent to OWTBs as a case study. The results demonstrate that the proposed damage index offers superior reliability in damage identification compared with traditional indicators when considering material aging. Moreover, the optimized stacking machine learning approach outperforms single machine learning methods, resulting in higher accuracy in damage identification.