APPLICATION OF ADAPTIVE FILTERING IN BEARING FAULT DETECTION IN WIND TURBINE GEAR TRANSMISSION SYSTEM

摘要

Wind turbines are developing and deploying fast, as wind power is becoming the world's fastest growing renewable energy source. In the future, reducing the operating cost of wind turbines is one of the critical issues with the growth of wind power. Maintenance of the wind turbine systems is high. Condition monitoring of transmission system of the wind turbines greatly reduce the maintenance cost, avoid catastrophic failure, and improve the reliability of the whole system. Bearing faults are one of the most common faults in wind turbines and bearings generate relatively weak signals which are usually buried in the background noise and vibration generated from other components. Also wind turbine transmission systems work under dynamic operating conditions. Thus developing advanced signal processing methods to effectively extract the bearing fault information is very important. In this paper, an adaptive filtering technique will be applied for bearing fault detection in wind turbine gear transmission systems. The periodic components are removed from the original vibration signal to enhance the bearing fault signal-to-noise ratio. Statistical features of the processed signal are extracted to quantify the bearing states. Simple linear classifier is trained to classify the healthy gearbox from the gearbox with bearing damage. Real wind turbine vibration signals were used to demonstrate the effectiveness of the presented method.