Research on data driven wind turbine blade structure health monitoring technology based on vibration analysis

Based on the CMS (Condition Monitoring System) data of wind turbines, this article conducts in-depth analysis of the frequency domain characteristics of wind turbine blades to diagnose whether wind turbine blades are damaged. Firstly, introduce the installation location of the sensor and the basic information of the collected data. Then describe in detail the basic process of frequency domain analysis, including condition screening, frequency domain transformation, frequency conversion analysis, and natural frequency analysis. By combining Fourier transform, three monitoring indicators are constructed, namely the effective value of the frequency band, the theoretical rate of change, and the relative rate of chang e. Data validation shows that based on these indicators, a statistical model can be established through comparison and statistical analysis to accurately evaluate the vibration characteristics of blades at different frequencies to identify whether the blad es have been damaged, with an accuracy rate of 95.1% and a recall rate of 100%.The research results indicate that the method proposed in this paper, based on Fourier transform and combined with the frequency domain characteristics of vibration data, constructs three monitoring indicators for the effective value of the first-order natural frequency of oscillation, theoretical change rate, and relative change rate to detect damage to wind turbine blades. This method can effectively diagnose whether wind turbine blades have been damaged and further improve the safety and reliability of wind turbine operations.