Diversity maximization-based transfer diagnosis approach of rotating machinery

Abstract

The existing transfer diagnosis methods based on entropy minimization are easy to lead to trivial solution. To solve this problem, a deep diversity maximization-based adversarial transfer diagnosis approach for rotating machinery is presented in this paper. Firstly, the deep convolution neural network is utilized as the feature encoder to learn the characteristics of vibration signals in different working conditions. The diversity maximization strategy is taken to balance the entropy minimization, so as to avoid trivial local minimum. The categories predicted by nontrivial domain adaptation method are more diverse. Moreover, the entropy is conducted to evaluate the uncertainty of the predicted result of the classifier. Using this deterministic strategy based on entropy to adjust the domain discriminator. The experimental study demonstrates the effectiveness of the developed method.

Remaining useful life prediction of ball screw based on integrating preload and precision

“…To further assess the accuracy and predictive capability of the three models, the Akaike information criterion (AIC) is introduced [38]. AIC, based on the maximum entropy method, enables the comparison of differences between models and can be expressed as equation (10). The candidate model with the smallest AIC value is selected for model choice.…”

“…The data-driven method is rooted in artificial intelligence technologies such as principal component analysis (PCA), partial least squares, and kernel PCA (KPCA), or in multivariate statistical techniques like deep learning, transfer learning, support vector machine, and artificial neural networks [7][8][9][10][11]. Contrasting with the physical-based approach, this methodology typically does not demand an in-depth comprehension of the intricate physics of the system.…”

An Adversarial Single-Domain Generalization Network for Fault Diagnosis of Wind Turbine Gearboxes

“…The batch size is uniformly set to 32. To eliminate the impact of randomness, the results are presented as the average of 20 trials [48].…”

A meta transfer learning method for gearbox fault diagnosis with limited data

“…Wu et al [25] illustrated a transfer diagnosis method using an adaptive Coral (correlation alignment) loss based attention mechanism and adversarial network. She et al [26] employed a adversarial transfer diagnosis scheme by using the diversity maximization strategy. Ren et al developed [27] a cross domain fault diagnosis model called dynamic balanced domain adversarial network The model used Wasserstein distance to measure distribution differences and was trained in an adversarial manner.…”