A Scalable Spark-Based Fault Diagnosis Platform for Gearbox Fault Diagnosis in Wind Farms

Gearbox faults in wind turbines are one of the most important reasons for the failure of these machines which lead to the longest downtime and maintenance cost. While much attention has been given to detect faults in these mechanical devices, real-time fault diagnosis for streaming vibration data from turbine gearboxes still remains an outstanding problem. Moreover, monitoring gearboxes in a wind farm with thousands of wind turbines requires massive computational power. In this paper, we propose a novel feature extraction algorithm to diagnose wind turbines' fault using vibration signal. We also implemented the whole system on an Apache Spark, a distributed framework for processing stream data. Using Spark clustering enables the fault diagnosis system to scale to large wind farms. The proposed algorithm has been tested by real-world wind turbine data under a different number of input sources, and an accuracy of 98.93% was obtained. Furthermore, a runtime analysis was done to evaluate the effect of parallelization using Spark stream processing.