Joint Application of VMD and IWOA-PNN for Gearbox Fault Classification via Current Signal
Jiesi Luo, Yuguang Chen, Qiaoyuan Huang, Shaohui Zhang, Xinbo Zhang
Abstract
In this article, a new method for fault diagnosis of wind turbine gearbox based on intelligent learning of current signal is proposed, which is a fusion method based on variational modal decomposition (VMD) and improved whale optimization algorithm (IWOA)-probabilistic neural network (PNN) classification. In view of the fundamental frequency component and noise interference of the current signal, VMD is used to decompose the current signal to get the fault-related intrinsic mode functions (IMFs). The energy entropy of IMFs and some time domain and frequency domain indexes are selected to form a feature dataset as the input of the IWOA-PNN classifier. In order to improve the classification effect of PNN, an IWOA based on the opposite-based learning (OBL) strategy and the crisscross optimization algorithm are designed to select the optimal smoothing factor, namely IWOA-PNN. The effectiveness and superiority of the proposed method have been demonstrated by both public data and self-test data.