Fault Diagnosis Method for Wind Turbine Gearboxes Based on IWOA-RF

A fault diagnosis method for wind turbine gearboxes based on undersampling, XGBoost feature selection, and improved whale optimization-random forest (IWOA-RF) was proposed the problem of high false negative positive rates in gearboxes. Normal samples raw data were subjected to undersampling first, various features labels provided with importance analysis by selection select higher label correlation. Two parameters random algorithm optimized via optimization create a fitness function rate (FNR) (FPR) as evaluation indexes. Then, minimum value within given scope found. The WOA controlled hyper-parameter ? optimize step size. This article uses variant form sigmoid alter change trend from linear decline rapid first then slow allow be optimized. In initial stage, larger size can make model progress target faster, while later stage optimization, smaller allows more accurately find function. Finally, two hyper-parameters, corresponding value, substituted into training. results showed that this paper significantly reduce compared other classification methods.