Fault Diagnosis via Kalman Filters and ANFIS Classifiers for a Wind Turbine

The operation mode of wind turbines becomes more challenging because they are exposed to environmental factors and operating conditions for extended periods. Optimizing their availability involves preventing faults effectively and early to avoid production losses. To preserve this critical power source, a fault detection and isolation strategy must be planned to manage all potential faults. Sensors, system components, and actuators of wind turbines are susceptible to various types of faults. This paper aims to propose an efficient fault detection and isolation system based on artificial intelligence techniques, such as ANFIS models and Kalman observers, for the wind turbine machine. An equivalent model is elaborated for residual generation and fault diagnosis in horizontal axis wind turbine parts. The proposed structure can generate effective residuals following the occurrence of faults and determine their type and location. The simulation results demonstrate the effectiveness of the proposed fault detection and isolation strategy for wind turbine machines.