Phase space reconstruction features for LSTM-based fault classification in gearboxes and roller bearings

Abstract

Phase space reconstruction (PSR) converts a vibration signal into a multidimensional trajectory encoding the nonlinear dynamics of rotating machinery, yet existing PSR-based deep learning approaches rasterize this trajectory into an image before classification, discarding its inherent temporal ordering. This paper proposes feeding two derivative-augmented PSR constructions—the Approximated Phase Space (APS) and the Classical Phase Space (CPS), directly as sequential inputs to a Long Short-Term Memory (LSTM) neural network, exploiting the structural compatibility between the column-sequential organization of the PSR matrix and LSTM’s temporal processing, and avoiding the image-conversion step used in CNN-based PSR approaches on the two datasets studied here. Each PSR matrix is augmented with the first and second derivatives computed via the Savitzky–Golay filter, extending the feature space to a derivative-augmented state representation that jointly encodes signal amplitude, rate of change, and curvature. Both constructions are validated through k-fold cross-validation on a 10-class spur gearbox dataset and a 7-class roller bearing dataset, both acquired under variable speed and load conditions. File-level classification via majority voting over feature pieces yields 99.7% accuracy for the gearbox and 99.7% for roller bearings using CPS features, and 94.6% and 99.0%, respectively, using APS features. CPS consistently outperforms APS by 9–14 percentage points at the piece level, attributable to its overlapping lag structure, while the majority voting substantially closes this gap at the file level. These results show that, on the two datasets evaluated, PSR matrices can be fed directly as sequential inputs to an LSTM without image conversion, yielding competitive fault classification accuracy, and that derivative-augmented phase space features provide a computationally efficient representation for fault diagnosis in rotating machinery.

Keywords

phase space reconstruction long short-term memory neural network rotating machinery fault diagnosis gearbox roller bearings vibration signal classification

Get full access to this article

View all access options for this article.

References

Cerrada

Sánchez

Pacheco

, et al. (2016) Hierarchical feature selection based on relative dependency for gear fault diagnosis. Applied Intelligence 44(3): 687–703. https://doi.org/10.1007/s10489-015-0725-3

Cerrada

Sánchez

, et al. (2018) A review on data-driven fault severity assessment in rolling bearings. Mechanical Systems and Signal Processing 99: 169–196. https://doi.org/10.1016/j.ymssp.2017.06.012

Chen

Zhang

, et al. (2024) A knowledge-based fault diagnosis method for rolling bearings without fault sample training. Proceedings of the Institution of Mechanical Engineers - Part C: Journal of Mechanical Engineering Science 238(20): 10253–10265. https://doi.org/10.1177/09544062241264267

Damou

Ratni

Benazzouz

(2024) Intelligent multi-fault identification and classification of defective bearings in gearbox. Advances in Mechanical Engineering 16(4): 16878132241246673. https://doi.org/10.1177/16878132241246673

Dang

Yang

Yuan

, et al. (2026) Fault diagnosis of wheelsets under non-periodic fragmented signals: a single-sensor approach based on phase space reconstruction. Measurement Science and Technology 37(8): 086111. https://doi.org/10.1088/1361-6501/ae46c0

Feeny

Lin

(2004) Fractional derivatives applied to phase-space reconstructions. Nonlinear Dynamics 38(1): 85–99. https://doi.org/10.1007/s11071-004-3748-6

Georgoulas

Loutas

Stylios

, et al. (2013) Bearing fault detection based on hybrid ensemble detector and empirical mode decomposition. Mechanical Systems and Signal Processing 41(1-2): 510–525. https://doi.org/10.1016/j.ymssp.2013.02.020

Hochreiter

Schmidhuber

(1997) Long short-term memory. Neural Computation 9(8): 1735–1780. https://doi.org/10.1162/neco.1997.9.8.1735

Kumar

Chauhan

Pandit

(2022) Time domain vibration analysis techniques for condition monitoring of rolling element bearing: a review. Materials Today: Proceedings 62: 6336–6340. https://doi.org/10.1016/j.matpr.2022.02.550

10.

Sánchez

Zurita

, et al. (2016) Fault diagnosis for rotating machinery using vibration measurement deep statistical feature learning. Sensors 16(6): 895. https://doi.org/10.3390/s16060895

11.

Liu

(2022) Intelligent fault diagnosis of rolling bearings under imbalanced data conditions using attention-based deep learning method. Measurement 189: 110500. https://doi.org/10.1016/j.measurement.2021.110500

12.

Medina

Macancela

Lucero

, et al. (2019a) Vibration signal analysis using symbolic dynamics for gearbox fault diagnosis. The International Journal of Advanced Manufacturing Technology 104(5-8): 2195–2214. https://doi.org/10.1007/s00170-019-03858-0

13.

Medina

Macancela

Lucero

, et al. (2019b) A LSTM neural network approach using vibration signals for classifying faults in a gearbox. In: 2019 International Conference on Sensing, Diagnostics, Prognostics, and Control (SDPC). IEEE, pp. 208–214. DOI: 10.1109/SDPC.2019.00045.

14.

Medina

Macancela

Lucero

, et al. (2022) Gear and bearing fault classification under different load and speed by using Poincaré plot features and SVM. Journal of Intelligent Manufacturing 33(4): 1031–1055. https://doi.org/10.1007/s10845-020-01712-9

15.

Halkon

, et al. (2023) Physics-informed residual network (piresnet) for rolling element bearing fault diagnostics. Mechanical Systems and Signal Processing 200: 110544. https://doi.org/10.1016/j.ymssp.2023.110544

16.

Packard

Crutchfield

Farmer

, et al. (1980) Geometry from a time series. Physical Review Letters 45(9): 712–716. https://doi.org/10.1103/physrevlett.45.712

17.

Peña

Cerrada

Medina

, et al. (2023) Poincaré plot features and statistical features from current and vibration signals for fault severity classification of helical gear tooth breaks. Journal of Computing and Information Science in Engineering 23(2): 021009. https://doi.org/10.1115/1.4054574

18.

Raubitzek

Neubauer

(2022) Reconstructed phase spaces and lstm neural network ensemble predictions. Engineering Proceedings 18(1): 40.

19.

Ren

, et al. (2025) Motor bearing fault diagnosis based on phase space reconstruction and Igam-Swin transformer. In: 2025 10th Asia Conference on Power and Electrical Engineering (ACPEE). IEEE, pp. 202–206.

20.

Saha

Hoque

Badihi

(2022) Development of intelligent fault diagnosis technique of rotary machine element bearing: a machine learning approach. Sensors 22(3): 1073. https://doi.org/10.3390/s22031073

21.

Savitzky

Golay

(1964) Smoothing and differentiation of data by simplified least squares procedures. Analytical chemistry 36(8): 1627–1639. https://doi.org/10.1021/ac60214a047

22.

Shahhosseini

Tien

D’Souza

(2023) Poincare maps: a modern systematic approach toward obtaining effective sections. Nonlinear Dynamics 111(1): 529–548. https://doi.org/10.1007/s11071-022-07864-y

23.

Tama

Vania

Lee

, et al. (2023) Recent advances in the application of deep learning for fault diagnosis of rotating machinery using vibration signals. Artificial Intelligence Review 56(5): 4667–4709. https://doi.org/10.1007/s10462-022-10293-3

24.

Wang

Ding

(2023) A new fault diagnosis of rolling bearing based on phase-space reconstruction and convolutional neural network. Industrial Lubrication & Tribology 75(8): 875–882. https://doi.org/10.1108/ILT-04-2023-0113

25.

Wang

Deng

, et al. (2019) Convolutional neural network based on spiral arrangement of features and its application in bearing fault diagnosis. IEEE Access 7: 64092–64100. https://doi.org/10.1109/ACCESS.2019.2916024

26.

Wang

Guo

, et al. (2020) Planetary-gearbox fault classification by convolutional neural network and recurrence plot. Applied Sciences 10(3): 932. https://doi.org/10.3390/app10030932

27.

Wang

, et al. (2025) Rolling bearing fault diagnosis based on multi-domain features and whale optimized support vector machine. Journal of Vibration and Control 31(5-6): 708–720. https://doi.org/10.1177/10775463241231344

28.

Zhang

Peng

, et al. (2026) Phase space reconstruction-based online roughness monitoring in robotic milling of aircraft skin edge. IEEE/ASME Transactions on Mechatronics 31(2): 1741–1752. Available at: https://doi.org/10.1109/TMECH.2025.3603285