Bearing remaining useful life prediction based on polynomial fitting and ABiLSTM-MLP

Résumé fourni par la source

Accurate remaining useful life (RUL) prediction of rolling bearings plays an important role in ensuring safe operation of machinery and reducing maintenance losses. To improve the accuracy of RUL prediction of rolling bearing, this paper proposes a RUL prediction method based on polynomial fitting and ABiLSTM-MLP. This method decomposes bearing signals by using discrete wavelet transform, extracts sensitive features from the obtained detailed components and approximate components, screens out time-domain features that can characterize the trend of bearing degradation, and constructs comprehensive performance degradation indicators by combining polynomial fitting. In addition to reducing the interference of signal noise components, a more representative degradation index is obtained. Based on the fusion of multi-layer perceptron with bidirectional long short-term memory networks, the model introduces an attention mechanism, which enables it to process time series data effectively and strengthens its ability to extract key information. Finally, the proposed method is validated by the XJTU-SY bearing dataset. Experimental results show that the proposed method can obtain better RUL prediction results.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.