This paper addresses the problems of traditional methods in the reliability assessment of automated electrical appliances, such as over-reliance on physical models, difficulty in fully utilizing multi-source monitoring data, and limited prediction accuracy. It proposes a research project based on big data intelligent algorithms. First, a multi-source heterogeneous data layer integrating operating parameters, condition monitoring, and maintenance records is constructed. Next, degradation features are extracted through time-domain, frequency-domain, and time-frequency-domain analysis, and key sensitive features are screened using methods such as ANOVA (Analysis of Variance). Then, an attention-enhanced LSTM (Long Short-Term Memory) network is introduced to learn performance degradation trajectories, and a stacking ensemble strategy is employed to combine the advantages of multiple models for remaining lifetime prediction. Finally, reliability quantification is achieved by combining survival analysis theory. Experimental results show that feature engineering improves the silhouette coefficient of the condition representation from 0.52 to 0.71, and the proposed ensemble model predicts a low RMSE of 21.1 hours, significantly outperforming traditional methods. This study confirms the effectiveness and superiority of data-driven intelligent algorithms in improving the accuracy of electrical appliance reliability assessment and lifetime prediction.
Hao et al. (Thu,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: