Naturally Damaged Wind Turbine Blade Bearing Fault Detection Using Novel Iterative Nonlinear Filter and Morphological Analysis

Wind turbine blade bearings are pivotal components to pitch blades, optimize electrical energy output and stop wind turbines for protection. Blade bearing failure can cause the turbine to lose control or even break down. However, due to the very slow rotation speeds (often less than 5 rpm) and limited rotation angles (less than 100o), blade bearings can only produce weak and limited operating condition data, which makes condition monitoring and fault diagnosis (CMFD) very challenging, in particular for naturally damaged conditions. In this paper, a naturally damaged large-scale blade bearing which was in operation on a real wind farm for over 15 years is investigated. An iterative nonlinear filter (INF) is proposed to remove heavy noise and extract weak fault vibration features. Then, the morphological transform-based envelope method is applied to diagnose the bearing fault in the frequency domain. The diagnostic results show that the proposed method can be an effective tool for diagnosing very slow speed blade bearings and is superior to conventional bearing fault diagnosis methods.