International Journal of Advanced Engineering Application

Deep Learning-Based Fault Detection in Three-Phase Induction Motors Using Vibration Signal Processing and 1D Convolutional Neural Networks

Author(s):Rajkumar Desai

Affiliation: Department of Electrical Engineering, Indian Institute of Technology Roorkee, Roorkee, Uttarakhand, India

Page No: 31-35

Volume issue & Publishing Year: Volume 3, Issue 7, 2026/07/05

Journal: International Journal of Advanced Engineering Application (IJAEA)

ISSN NO: 3048-6807

DOI:

Download PDF Cite this article

Abstract:
Induction motors constitute over 65% of industrial electrical energy consumption globally, and unplanned motor failures account for significant production losses and maintenance costs across manufacturing and process industries in India. Bearing faults, broken rotor bars (BRB), and stator winding insulation failures together represent more than 75% of all induction motor failures. Conventional vibration-based fault detection methods relying on hand-crafted time-frequency features extracted by domain experts exhibit limited generalisation across motor sizes, load conditions, and installation environments. This study proposes a one-dimensional Convolutional Neural Network (1D-CNN) framework that operates directly on raw triaxial vibration signals sampled at 12 kHz, eliminating the need for manual feature engineering. A dataset of 6,000 labelled vibration signal segments — 1,500 per class (Healthy, Broken Rotor Bar, Inner Raceway Fault, Outer Raceway Fault) — was acquired from a 7.5 kW, 415 V, 50 Hz induction motor testbed under variable load conditions (0%, 25%, 50%, 75%, 100% rated load). The proposed 1D-CNN architecture with three convolutional blocks, global average pooling, and two dense layers achieves 97.1% classification accuracy and 96.7% macro F1-score on the held-out test set, outperforming SVM (87.4%), Random Forest (89.2%), and LSTM (92.8%) baselines. Robustness evaluation under additive Gaussian noise at SNR levels from −10 dB to +20 dB confirms the model's utility in real industrial acoustic environments. The proposed approach achieves convergence in under 50 epochs and inference latency of 2.3 ms per segment on an ARM Cortex-M7 embedded processor, confirming embedded deployment feasibility for edge-based condition monitoring.

Keywords: induction motor, fault detection, 1D-CNN, vibration signal, broken rotor bar, bearing fault, deep learning, condition monitoring, edge inference

Reference:

• [1] Abid, A., Khan, M. T., & Iqbal, J. (2021). A review on fault detection and diagnosis techniques: basics and beyond. Artificial Intelligence Review, 54(5), 3639-3664.
• [2] Agarwal, D., & Goel, P. (2022). Deep learning-based bearing fault detection in induction motors using vibration signals. Journal of Vibration Engineering and Technologies, 10(4), 1523–1535.
• [3] Benbouzid, M. E. H. (2000). A review of induction motors signature analysis as a medium for faults detection. IEEE Transactions on Industrial Electronics, 47(5), 984-993.
• [4] Chen, Z., & Gryllias, K. (2019). Intelligent fault diagnosis for rotary machinery using transferable convolutional neural network. IEEE Transactions on Industrial Informatics, 16(1), 339-349.
• [5] Devarakonda, S., & Reddy, K. (2023). Motor current signature analysis for induction motor fault identification in Indian process industries. IE(I) Journal — Electrical Engineering, 104(2), 88-96.
• [6] Gangsar, P., & Tiwari, R. (2020). Signal based condition monitoring techniques for fault detection and diagnosis of induction motors: a state-of-the-art review. Mechanical Systems and Signal Processing, 144, 106908.
• [7] Ince, T., Kiranyaz, S., Eren, L., Askar, M., & Gabbouj, M. (2016). Real-time motor fault detection by 1D convolutional neural networks. IEEE Transactions on Industrial Electronics, 63(11), 7067-7075.
• [8] Kiranyaz, S., Avci, O., Abdeljaber, O., Ince, T., Gabbouj, M., & Inman, D. J. (2021). 1D convolutional neural networks and applications: A survey. Mechanical Systems and Signal Processing, 151, 107398.
• [9] Kumar, R., & Singh, M. (2021). Bearing fault diagnosis in rotating machinery using deep learning. Measurement, 172, 108855.
• [10] Lei, Y., Yang, B., Jiang, X., Jia, F., Li, N., & Nandi, A. K. (2020). Applications of machine learning to machine fault diagnosis: A review and roadmap. Mechanical Systems and Signal Processing, 138, 106587.
• [11] Patro, S. G. K., & Sahu, K. K. (2015). Normalization: A preprocessing stage. arXiv preprint arXiv:1503.06462.
• [12] Rathore, A., & Paliwal, M. (2022). Edge-deployed deep learning inference for industrial condition monitoring: hardware-software co-design. IEEE Embedded Systems Letters, 14(3), 119-122.
• [13] Selesnick, I. W., & Graber, H. L. (2021). Transient artifact reduction in vibration signals using sparsity-regularised signal decomposition. Signal Processing, 179, 107803.