Condition Monitoring for the Roller Bearings of Wind Turbines under Variable Working Conditions Based on the Fisher Score and Permutation Entropy

摘要: Condition monitoring is used to assess the reliability and equipment efficiency of wind turbines. Feature extraction an essential preprocessing step achieve a high level performance in condition monitoring. However, fluctuating conditions turbines usually cause sudden variations monitored features, which may lead inaccurate prediction maintenance schedule. In this scenario, article proposed novel methodology detect multiple levels faults rolling bearings variable operating conditions. First, signal decomposition was carried out by variational mode (VMD). Second, statistical features were calculated extracted time domain. Meanwhile, permutation entropy analysis conducted estimate complexity vibrational series. Next, feature selection techniques applied improved identification accuracy reduce computational burden. Finally, ranked vectors fed into machine learning algorithms for classification bearing defect status. particular, method performed over wide range working regions simulate operational Comprehensive experimental investigations employed evaluate effectiveness method.