Stacking ensemble learning models to improve fault-diagnosis accuracy

Abstract

As the cornerstone of modern industrial production, the reliable operation of complex mechanical equipment is crucial for ensuring operational stability, enhancing production efficiency, and preventing safety incidents. However, due to the intricate structures, multi-source coupled fault modes, and prominent nonlinear dynamic evolution characteristics, traditional diagnostic models often fail to simultaneously address high-dimensional feature processing and complex nonlinear relationships, which limits their diagnostic accuracy. To address the high-accuracy diagnostic requirements in industrial applications, this paper proposes a fault-diagnosis method for complex machinery based on Stacking ensemble learning. The method integrates Random Forest (RF), XGBoost, Support Vector Machine (SVM), and Backpropagation (BP) neural networks as heterogeneous base learners, with a BP neural network employed as the meta-learner to construct a layered fusion framework that effectively combines the strengths of each model. In addition, feature selection is performed using time-weighted maximum mutual information coefficients in combination with recursive feature-elimination cross-validation to identify the optimal feature subset. Furthermore, a constraint-based Bayesian optimization approach, tailored for industrial deployment, is used for automatic hyperparameter tuning of the base learners. The proposed method is applied to a lithium battery electrode double-roll pressing machine, where comprehensive data completion and validation experiments are conducted on real-world data containing 11 fault categories. The results show that the proposed method achieves fault classification and severity assessment with a diagnostic accuracy of 96.67%, significantly outperforming individual models and conventional fusion strategies. This study provides an effective technical pathway for the precise and efficient intelligent fault diagnosis of complex mechanical equipment.

Keywords

complex mechanical equipment fault diagnosis ensemble learning optimization Stacking model base learners meta-learners

Get full access to this article

View all access options for this article.

References

Abro

Siddique

Talpur

MSH

, et al. (2023) A combined approach of base and meta learners for hybrid system. Turkish Journal of Engineering 7(1): 25–32.

Bahad

Saxena

(2020) Study of adaboost and gradient boosting algorithms for predictive analytics. In: International conference on intelligent computing and smart communication 2019: Proceedings of ICSC 2019, Tehri, India, 20–21 April 2019, pp.235–244. Springer.

Bischl

Binder

Lang

, et al. (2023) Hyperparameter optimization: Foundations, algorithms, best practices, and open challenges. Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery 13(2): e1484.

Chen

Zheng

(2023) Bearing fault diagnosis based on feature screening and integrated learning. Journal of Shanghai Electric Power University 39(5): 489–495.

Chen

Zhao

Han

(2025) Uncertainty-guided Bayesian active learning for cost-effective fault diagnosis with minimal labeled data. Engineering Applications of Artificial Intelligence 160: 111945.

Darst

Malecki

Engelman

(2018) Using recursive feature elimination in random forest to account for correlated variables in high dimensional data. BMC Genetics 19: 1–6.

Deng

Shan

, et al. (2019) Processing methods of missing data and its development trend. Statistics and Decision Making 35(23): 28–34.

Divi

Priyadharsini

(2015) Design of Monitoring and Fault Diagnosis System in Wind Turbine Based on CAN Bus. International Journal of Electrical and Electronics Research 3(4): 79–81.

Dong

Cao

, et al. (2020) A survey on ensemble learning. Frontiers of Computer Science 14: 241–258.

10.

Fan

Wang

Gan

, et al. (2020) Research on fault diagnosis technology of launch vehicle based on fault tree. Measurement and Control Technology 39(10): 19–23+31.

11.

Guo

Yang

Hua

, et al. (2021a) Coupling fault diagnosis of wind turbine gearbox based on multitask parallel convolutional neural networks with overall information. Renewable Energy 178: 639–650.

12.

Guo

Ding

Zheng

, et al. (2021b) Research on coupled fault separation method of hydraulic pump with new structural elements. Machine Tools and Hydraulics 49(5): 162–167.

13.

Miao

Shen

(2023a) Prediction of combat effectiveness of weapon and equipment system based on improved Stacking integrated learning method. Journal of Military Engineering 44(11): 3455–3464.

14.

Jiao

Liu

, et al. (2025) Trustworthy Bayesian deep learning framework for uncertainty quantification and confidence calibration: Application in machinery fault diagnosis. Reliability Engineering & System Safety 255: 110657.

15.

Chen

Jiao

(2023b) Oil content prediction of near-infrared spectral oil shale based on Stacking integrated learning. Spectroscopy and Spectral Analysis 43(4): 1030–1036.

16.

Nie

, et al. (2019) A review of feature selection methods. Computer Engineering and Applications 55(24): 10–19.

17.

Liu

Chen

Zhao

(2023) Integrated learning method based on attention mechanism and its research in industrial fault diagnosis. Journal of Electronics 51(3): 579–587.

18.

Mienye

Sun

(2022) A survey of ensemble learning: Concepts, algorithms, applications, and prospects. IEEE Access 10: 99129–99149.

19.

Misra

Yadav

(2020) Improving the classification accuracy using recursive feature elimination with cross-validation. International Journal on Emerging Technologies 11(3): 659–665.

20.

Mustaqim

Adi

Pristyanto

, et al. (2021) The effect of recursive feature elimination with cross-validation (RFECV) feature selection algorithm toward classifier performance on credit card fraud detection. In: 2021 International conference on artificial intelligence and computer science technology (ICAICST), Yogyakarta, Indonesia, 29–30 June 2021, pp.270–275. IEEE.

21.

Probst

Boulesteix

Bischl

(2019) Tunability: Importance of hyperparameters of machine learning algorithms. The Journal of Machine Learning Research 20(1): 1934–1965.

22.

Soltanali

Khojastehpour

Farinha

, et al. (2021) An integrated fuzzy fault tree model with Bayesian Network-Based maintenance optimization of complex equipment in automotive manufacturing. Energies 14(22): 7758.

23.

Sun

Wen

, et al. (2021) Assessment of landslide susceptibility mapping based on Bayesian hyperparameter optimization: A comparison between logistic regression and random forest. Engineering Geology 281: 105972.

24.

Wang

Xing

Wang

, et al. (2019) Maximal information coefficient and support vector regression based nonlinear feature selection and QSAR modeling on toxicity of alcohol compounds to Tadpoles of Rana temporaria. Journal of the Brazilian Chemical Society 30: 279–285.

25.

, et al. (2023) Rock compressive strength prediction based on integrated learning and Bayesian optimization. Earth Science 48(5): 1686–1695.

26.

Xiong

Guo

(2021) A review of research on missing data processing methods. Computer Engineering and Applications 57(14): 27–38.

27.

Yang

(2018) Integrated learning methods: A research review. Journal of Yunnan University (Natural Science Edition) 40(6): 1082–1092.

28.

Wang

, et al. (2019) Aero-engine fault diagnosis method based on integrated learning. Journal of Civil Aviation University of China 37(2): 29–3342.

29.

Yang

Liu

Zhang

, et al. (2022) Landslide susceptibility assessment based on Bayesian optimization machine learning hyperparameters. Geoscience and Technology Bulletin 41(2): 228–238.

30.

Yang

Shami

(2020) On hyperparameter optimization of machine learning algorithms: Theory and practice. Neurocomputing 415: 295–316.

31.

Zheng

Cheng

(2019) Churn user prediction method based on multi-model fusion. Computer Engineering and Science 41(11): 2027–2032.

32.

Zhang

Dou

(2021a) A new method for accuracy evaluation of multi-classification tasks based on confusion matrix. Computer Engineering and Science 43(11): 1910–1919.

33.

Zhang

Guo

Liu

, et al. (2021b) Multi-model integrated short-term load forecasting based on maximum mutual information coefficient and wavelet decomposition. Computer Applications and Software 38(5): 82–87.

34.

Zheng

Wang

Yao

, et al. (2023) Dynamic mechanical strength prediction of BFRC based on stacking ensemble learning and genetic algorithm optimization. Buildings 13(5): 1155.

35.

Zhou

Lei

(2021) Fault diagnosis method of the construction machinery hydraulic system based on artificial intelligence dynamic monitoring. Mobile Information Systems 2021: 1–10.