کنترل‌کننده تحمل‌پذیر عیب بلادرنگ موتورهای القایی با استفاده از شبکه عصبی عمیق CNN-LSTM بهینه‌شده مبتنی بر داده‌های ارتعاش-آکوستیک و الگوریتم کنترل پیشنهادی

نوع مقاله : مقاله علمی

نویسندگان
عبداله صفری دهنوی 1
وحید صفری دهنوی 2
1 مربی، گروه مهندسی کشاورزی، دانشگاه ملی مهارت، تهران، ایران
2 دانشجوی دکتری، گروه مهندسی برق کنترل، دانشگاه صنعتی امیرکبیر، تهران، ایران
10.22034/stme.2025.533639.1144
چکیده
تعمیرات پیشگیرانه و کنترل سرعت ماشین‌های دوار به دلیل کاهش هزینه‌های نگهداری، بهبود کارایی و قابلیت اطمینان در صنایع مدرن حیاتی هستند. این مقاله روشی برای تشخیص عیب و کنترل سرعت موتورهای القایی ارائه می‌دهد. روش پیشنهادی از دو الگوریتم برای تشخیص عیب استفاده می‌کند؛ در الگوریتم اول، از یک روش نوین برای استخراج ویژگی‌های آماری، فرکانسی و انرژی و همچنین از الگوریتم‌های یادگیری ماشین برای کلاس‌بندی عیب استفاده می‌شود. اگر دقت تشخیص عیب الگوریتم اول مناسب نباشد، از الگوریتم دوم استفاده می‌شود. این الگوریتم از شبکه عصبی کانولوشنی (CNN) و حافظه کوتاه‌مدت بلند (LSTM) بهینه‌شده استفاده می‌کند. در این شبکه، استخراج و انتخاب ویژگی به‌طور خودکار انجام می‌شود. ما یک الگوریتم جدید برای بهینه‌سازی لایه‌ها و پارامترها از طریق الگوریتم تطبیق یافته کلونی مورچگان و توابع احتمال ارائه دادیم. روش‌های پیشنهادی روی سه مجموعه داده آزمایش‌شده‌اند. در مجموعه داده اول که در این پژوهش جمع‌آوری شده است، تمرکز بر حالت سالم، عیب یاتاقان و عدم کوپل صحیح هست. در مجموعه داده دوم، بر روی عیوب حلقه‌های داخلی و خارجی یاتاقان تمرکز می‌کنیم. در مجموعه داده سوم، هشت حالت شامل هفت عیب موجود در استاتور و روتور و یک حالت سالم، در نظر گرفته شده است و از داده‌های ارتعاش و آکوستیک برای تشخیص عیب استفاده شده است. سپس، یک الگوریتم کنترلی، شامل یک کنترل‌کننده کنترل بهینه مرتبه کسری برای کنترل سرعت موتور به‌ویژه هنگامی‌که تحت عیب است، ارائه می‌شود. نتایج نشان می‌دهد که الگوریتم اول برای مجموعه داده اول و دوم و الگوریتم دوم برای مجموعه داده سوم مناسب بود.
کلیدواژه‌ها
تعمیرات پیشگیرانه
کنترل سرعت
یادگیری ماشین
شبکه عصبی عمیق
داده‌های ارتعاش و آکوستیک
موضوعات

عنوان مقاله English

Real-time fault-tolerant controller of induction motors using optimized CNN-LSTM deep neural network based on vibration-acoustic data and proposed control algorithm

نویسندگان English

Abdolah Safari Dehnavi 1
Vahid Safari Dehnavi 2
1 Instructor, Department of Agriculture Engineering, Technical and Vocational University (TVU), Tehran, Iran
2 Ph.D. Student, Department of Electrical Engineering, Amirkabir University of Technology (Tehran Polytechnic), Tehran, Iran
چکیده English

Predictive maintenance and speed control of rotating machines are crucial in modern industries due to the reduction of maintenance costs and the improvement of efficiency and reliability. This paper presents a method for fault diagnosis and speed control of induction motors. The proposed method employs two algorithms for fault diagnosis. In the first algorithm, a novel method for extracting statistical, frequency, and energy features is utilized, along with machine learning algorithms for fault classification. If the fault detection accuracy of the first algorithm is not suitable, the second algorithm is used. This algorithm utilizes a convolutional neural network (CNN) and an optimized long short-term memory (LSTM) model. In this network, feature extraction and selection are performed automatically. We present a new algorithm for optimizing layers and parameters through an adaptive ant colony algorithm and likelihood functions. The proposed methods are tested on three datasets. In the first dataset collected for this research, the focus is on the healthy state, bearing fault, and uncoupled condition. In the second dataset, we focus on the defects of the inner and outer rings of the bearing. In the third dataset, eight cases are considered, including seven cases with defects in the stator and rotor, and one healthy case. Vibration and acoustic data are used for fault diagnosis. Then, a control algorithm, including a fractional-order optimal control controller, is presented to control the motor speed, especially when it is under fault conditions. The results indicate that the first algorithm is suitable for the first and second datasets, while the second algorithm is more suitable for the third dataset.

کلیدواژه‌ها English

Preventive maintenance
speed control
machine learning
deep neural network
vibration and acoustic data

اصل مقاله

[1] V. S. Dehnavi and M. Shafiee, “Fault diagnosis of induction motors using novel measurement techniques and data fusion,” Measurement, Vol. 256, Art. No. 118135, 2025, doi: 10.1016/j.measurement.2025.118135.
[2] K. S. Krishna Veni and N. S. Kumar, “Diagnosis of bearing fault in induction motor using Bayesian optimization-based ensemble classifier,” Electrical Engineering, Vol. 106, pp. 1895–1905, 2023, doi: 10.1007/s00202-023-02040-w.
[3] A. Almounajjed, A. K. Sahoo, and M. K. Kumar, “Condition monitoring and fault detection of induction motor based on wavelet denoising with ensemble learning,” Electrical Engineering, Vol. 104, pp. 2859–2877, 2022, doi: 10.1007/s00202-022-01523-6.
[4] O. AlShorman et al., “Sounds and acoustic emission-based early fault diagnosis of induction motor: A review study,” Advances in Mechanical Engineering, Vol. 13, Art. No. 168781402199691, 2021, doi: 10.1177/1687814021996915.
[5] V. S. Dehnavi and M. Shafiee, “Inner and outer bearing fault diagnosis of electrical motors using a proposed algorithm and vibration signals,” in Proc. 14th Int. Conf. Information and Knowledge Technology (IKT), 2023, pp. 175–180, doi: 10.1109/IKT62039.2023.10433018.
[6] A. Glowacz, “Thermographic fault diagnosis of electrical faults of commutator and induction motors,” Engineering Applications of Artificial Intelligence, Vol. 121, Art. No. 105962, 2023, doi: 10.1016/j.engappai.2023.105962.
[7] L. Sheng et al., “Research on gear crack fault diagnosis model based on permanent magnet motor current signal,” ISA Transactions, Vol. 135, pp. 188–198, 2023, doi: 10.1016/j.isatra.2022.10.001.
[8] J. J. Saucedo-Dorantes et al., “Automatic methodology for multiple fault detection in induction motor under periodic low-frequency fluctuating load based on stray flux signals,” IEEE Transactions on Energy Conversion, Vol. 38, No. 4, pp. 2744–2753, 2023, doi: 10.1109/TEC.2023.3294392.
[9] J.-G. Jang et al., “Vibration data feature extraction and deep learning-based preprocessing method for highly accurate motor fault diagnosis,” Journal of Computational Design and Engineering, Vol. 10, pp. 204–220, 2022, doi: 10.1093/jcde/qwac128.
[10] A. Choudhary et al., “Multi-input CNN based vibro-acoustic fusion for accurate fault diagnosis of induction motor,” Engineering Applications of Artificial Intelligence, Vol. 120, Art. No. 105872, 2023, doi: 10.1016/j.engappai.2023.105872.
[11] R. Kashfi and S. H. Karimizadeh, “Fault diagnosis methods for induction motors using motor current signal analysis,” in Proc. 7th Int. Conf. Electrical, Computer and Mechanical Engineering, Tehran, Iran, 2021 (in Persian).
[12] B. Noori and M. Ojaghi, “Application of motor current signal analysis (MCSA) for stator core insulation fault diagnosis in three-phase induction motors,” in Proc. 8th Int. Conf. Electrical, Computer, Mechanical and Artificial Intelligence Engineering, Mashhad, Iran, 2024 (in Persian).
[13] M.-Q. Tran et al., “Effective fault diagnosis based on wavelet and convolutional attention neural network for induction motors,” IEEE Transactions on Instrumentation and Measurement, Vol. 71, pp. 1–13, 2022, doi: 10.1109/TIM.2021.3139706.
[14] H. Li et al., “A normalized frequency-domain energy operator for broken rotor bar fault diagnosis,” IEEE Transactions on Instrumentation and Measurement, Vol. 70, pp. 1–10, 2021, doi: 10.1109/TIM.2020.3009011.
[15] N. Jia et al., “Intelligent fault diagnosis of rotating machines based on wavelet time-frequency diagram and optimized stacked denoising auto-encoder,” IEEE Sensors Journal, Vol. 22, pp. 17139–17150, 2022, doi: 10.1109/JSEN.2022.3193943.
[16] P. Gangsar and R. Tiwari, “Signal-based condition monitoring techniques for fault detection and diagnosis of induction motors: A state-of-the-art review,” Mechanical Systems and Signal Processing, Vol. 144, Art. No. 106908, 2020, doi: 10.1016/j.ymssp.2020.106908.
[17] H. Tao et al., “Unsupervised cross-domain rolling bearing fault diagnosis based on time-frequency information fusion,” Journal of the Franklin Institute, Vol. 360, pp. 1454–1477, 2023, doi: 10.1016/j.jfranklin.2022.11.004.
[18] V. Safari Dehnavi and M. Shafiee, “Data-driven control framework using fractional order singular optimal control and optimized metaheuristic algorithms,” Computers & Electrical Engineering, Vol. 120, Art. No. 109728, 2024, doi: 10.1016/j.compeleceng.2024.109728.
[19] H. Wang, J. Zheng, and J. Xiang, “Online bearing fault diagnosis using numerical simulation models and machine learning classifications,” Reliability Engineering & System Safety, Vol. 234, Art. No. 109142, 2023, doi: 10.1016/j.ress.2023.109142.
[20] R. R. Shubita et al., “Fault detection in rotating machinery based on sound signal using edge machine learning,” IEEE Access, Vol. 11, pp. 6665–6672, 2023, doi: 10.1109/ACCESS.2023.3237074.
[21] R. Rajabioun et al., “Distributed bearing fault classification of induction motors using 2-D deep learning model,” IEEE Journal of Emerging and Selected Topics in Industrial Electronics, Vol. 5, pp. 115–125, 2024, doi: 10.1109/JESTIE.2023.3323253.
[22] S. Ayankoso et al., “Multisensory data-based fault diagnosis of induction motors using 1D and 2D convolutional neural networks,” Mechanisms and Machine Science, pp. 1125–1135, 2024, doi: 10.1007/978-3-031-49421-5_92.
[23] E. Sonmez et al., “A new deep learning model combining CNN for engine fault diagnosis,” Journal of the Brazilian Society of Mechanical Sciences and Engineering, 2023, doi: 10.1007/s40430-023-04537-8.
[24] P. Borghesani et al., “A Fourier-based explanation of 1D-CNNs for machine condition monitoring applications,” Mechanical Systems and Signal Processing, Vol. 205, Art. No. 110865, 2023, doi: 10.1016/j.ymssp.2023.110865.
[25] C. Celebioglu et al., “Smartphone-based bearing fault diagnosis in rotating machinery using audio data and 1D convolutional neural networks,” in Proc. Int. Conf. Computer Systems and Technologies, 2024, pp. 149–154, doi: 10.1145/3674912.3674918.
[26] S. Ippili et al., “Deep learning-based mechanical fault detection and diagnosis of electric motors using directional characteristics of acoustic signals,” Noise Control Engineering Journal, Vol. 71, pp. 384–389, 2023, doi: 10.3397/1/377132.
[27] Y. Alkhanafseh et al., “Advanced dual RNN architecture for electrical motor fault classification,” IEEE Access, Vol. 12, pp. 2965–2976, 2024, doi: 10.1109/ACCESS.2023.3344676.
[28] A. Balamurugan et al., “Fault diagnosis of three-phase induction motor using a hybrid ELSE-RNN technique,” IETE Journal of Research, pp. 1–10, 2024, doi: 10.1080/03772063.2024.2315199.
[29] C.-S. Tu et al., “An audio-based motor-fault diagnosis system with SOM-LSTM,” Applied Sciences, Vol. 14, Art. No. 8229, 2024, doi: 10.3390/app14188229.
[30] J. Chuya-Sumba et al., “Deep-learning method based on 1D convolutional neural network for intelligent fault diagnosis of rotating machines,” Applied Sciences, Vol. 12, No. 4, Art. No. 2158, 2022, doi: 10.3390/app12042158.
[31] S. Gao et al., “Bearing fault diagnosis based on adaptive convolutional neural network with Nesterov momentum,” IEEE Sensors Journal, Vol. 21, No. 7, pp. 9268–9276, 2021, doi: 10.1109/JSEN.2021.3050461.
[32] Z. Zhu et al., “A review of the application of deep learning in intelligent fault diagnosis of rotating machinery,” Measurement, Vol. 206, Art. No. 112346, 2023, doi: 10.1016/j.measurement.2022.112346.
[33] F. Wang et al., “Cascade convolutional neural network with progressive optimization for motor fault diagnosis under nonstationary conditions,” IEEE Transactions on Industrial Informatics, Vol. 17, No. 4, pp. 2511–2521, 2020, doi: https://doi.org/10.1109/TII.2020.3003353.
[34] P. Chen et al., “An automatic speed adaption neural network model for planetary gearbox fault diagnosis,” Measurement, Vol. 171, Art. No. 108784, 2021, doi: 10.1016/j.measurement.2020.108784.
[35] K. Sharma et al., “Intelligent fault diagnosis of bearings based on convolutional neural network using infrared thermography,” Proc. IMechE, Part J: Journal of Engineering Tribology, Vol. 236, No. 12, pp. 2439–2446, 2022, doi: 10.1177/13506501221082746.
[36] X. Zhao et al., “Intelligent fault diagnosis of gearbox under variable working conditions with adaptive intraclass and interclass convolutional neural network,” IEEE Transactions on Neural Networks and Learning Systems, Vol. 34, No. 9, pp. 6339–6353, 2022, doi: 10.1109/TNNLS.2021.3135877.
[37] F. Pourdadashi Komachali et al., “Design of unknown input fractional order proportional–integral observer for fractional order singular systems with application to actuator fault diagnosis,” IET Control Theory & Applications, Vol. 13, No. 14, pp. 2163–2172, 2019, doi: 10.1049/iet-cta.2018.5712.
[38] V. S. Dehnavi and M. Shafiee, “LQR for generalized systems using metaheuristic algorithms based on disturbance observer,” in Proc. 28th Iranian Conf. Electrical Engineering (ICEE), 2020, pp. 1–5, doi: 10.1109/ICEE50131.2020.9260723.
[39] A. Safari Dehnavi and V. Safari Dehnavi, “Hardware-software cyber security platform for data protection in smart agricultural wireless sensor network with signal processing capability,” Karafan Quarterly Scientific Journal, in press, 2025, doi: 10.48301/kssa.2024.476579.2982.
[40] W. Jung et al., “Vibration, acoustic, temperature, and motor current dataset of rotating machine under varying load conditions for fault diagnosis,” Mendeley Data, Vol. 6, 2023, doi: 10.17632/ZTMF3M7H5X.6.
[41] M. Sehri and P. Dumond, “University of Ottawa electric motor dataset – vibration and acoustic faults under constant and variable speed conditions (UOEMD-VAFCVS),” Mendeley Data, Vol. 1, 2023, doi: 10.17632/MSXS4VJ48G.1.
[42] S. Adigintla and M. V. Aware, “Design and analysis of a speed controller for fractional-order-modeled voltage-source-inverter-fed induction motor drive,” International Journal of Circuit Theory and Applications, Vol. 50, pp. 2378–2397, 2022, doi: 10.1002/cta.3290.
[43] M. Ertargin et al., “Mechanical and electrical faults detection in induction motor across multiple sensors with CNN-LSTM deep learning model,” Electrical Engineering, 2024, doi: 10.1007/s00202-024-02420-w.
[44] Y. Li et al., “Semi-supervised meta-path space extended graph convolution network for intelligent fault diagnosis of rotating machinery under time-varying speeds,” Reliability Engineering & System Safety, Vol. 251, Art. No. 110363, 2024.

  • تاریخ دریافت 20 تیر 1404
  • تاریخ بازنگری 28 مرداد 1404
  • تاریخ پذیرش 31 شهریور 1404
  • تاریخ اولین انتشار 31 شهریور 1404
  • تاریخ انتشار 01 بهمن 1404