Fault Prediction Method for Equipment Based on Stacking
The continuous and stable operation of machinery and equipment was crucial for ensu-ring production,and fault prediction technology was key to maintaining continuous operation.To further enhance the effectiveness of equipment fault prediction,a fault prediction method for machinery based on a Stacking ensemble learning model,called the TXL-RF model,was pro-posed.This model improved the accuracy of machinery fault prediction by integrating the ad-vantages of three advanced algorithms:TabNet,XGBoost,and LightGBM.The TXL-RF model independently trained the three base learners and used their prediction results as feature inputs to a random forest model for comprehensive fault risk assessment.The results showed that the TXL-RF model achieved an F1 score of 0.889 and an accuracy of 99.38％.It not only main-tained accuracy comparable to the base learners but also achieved a significant improvement in the F1 score,demonstrating the model's advantage in comprehensive prediction performance and the effectiveness of the ensemble strategy.
NETL
NSTL
万方数据
