SASLN: Signals Augmented Self-Taught Learning Networks for Mechanical Fault Diagnosis Under Small Sample Condition

The implementation of condition monitoring and fault diagnosis is of special importance for ensuring wind turbine (WT) operation safely and stably. In practice, however, the fault data of WT are limited, which makes it hard to identify faults of WT accurately using the existing intelligent diagnosis methods. To address this, signals augmented self-taught learning network (SASLN) is proposed for the fault diagnosis of the generator, which is one of the most important parts in WT. In SASLN, fault signal samples are generated by the Wasserstein distance guided generative adversarial networks to expand the limited training data set. The sufficient generated signal samples are used to pretrain the self-taught learning network (SIN) to enhance the generalization ability of SLN. Then, the weights of SIN are fine-tuned using a small number of real signal samples for accurate fault classification. The effectiveness of SASLN is verified by two bearing vibration data sets. The results show that SASLN can achieve fairly high fault classification accuracy using small training samples. Besides, SASLN has good robustness in noisy working environment and can also identify faults even in variable loads and variable rotating speeds cases, which makes it meaningful for decreasing the running costs and improving the maintenance management of WT.

Automated summary

The SASLN utilizes generative adversarial networks to expand limited training dataset for accurate fault diagnosis of generator in wind turbines. The network is pretrained with generated signal samples to enhance generalization ability before fine-tuning with real samples for accurate fault classification.