Transient stability assessment model based on stacked sparse denoising auto-encodern

With its good performance, the deep learning model is introduced into the transient stability assessment of power system. However, when applying online, the anti-noise ability and generalization ability of the models must be paid attention to. A deep learning model based on stacked sparse denoising auto-encoder (SSDAE) is proposed for real-time transient stability assessment. The corrupted data is generated by adding white Gaussian noise to the original input data. Then, the high-order features are extracted from the corrupted data. Finally, the original data is reconstructed by using the high-order features. The anti-noise ability is greatly improved after training. Besides, the weights and the degree of activation of the hidden layer neurons are suppressed to achieve the sparseness of the model, which can improve the generalization ability of the model. The simulation results show that the SSDAE based model has good anti-noise ability and generalization ability compared to other machine learning algorithms.

Abstract licence: CC BY-NC

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.