review meeting ppt updated- july 2022.pptx

Review meeting

1. Fault Prediction of Induction Motor using Soft Computing Techniques BY Mr. Jaheer .H. Patel Registration No.2018792110 Supervisor name: Dr.S.Barani Date: - 30/07/2022

2. CONTENTS Motivation Objectives Types of Faults Literature Survey Experimental Work for Vibration based fault detection Conditions Consider for Experiment Experiment validation Experiment validation

3. MOTIVATION Induction motor is an alternating current electric motor that is commonly used in industrial and commercial application . It is find that almost 80% of motors used in industry are Induction motor for the transformation of electrical energy to mechanical energy . Induction Motor are more popular than other motor because of economical, small size, ruggedness, reliability, low maintenance and low operation cost . All though Induction motors are very reliable, but they are exposed to environmental, duty and installation problems . These problem cause motor subjected to various types of failures and hence its life time become less. If detection of motor faults is not done at primary stage then maintenance is done before machine stop working .

4. OBJECTIVES The Objective of this research is to detect faults of Induction motor at primary stage hence maintenance is possible before machine stop working. To achieve this objective , develop an effective fault prediction method using machine learning method and vibration data . Vibration data is used because that help to detect all mechanical fault as well as electrical faults of Induction Motor.

5. Types of Faults Fig.1 Classification of Faults [1]

6. Experimental Work for Vibration based fault detection Fig2. Interfacing of MPU6050 with Arduino Uno

7. Fig3. Attachment of sensor MPU6050 with 2 HP Induction Motor

8. Conditions Consider for Experiment No load Normal Bearing Half load Normal Bearing Full load Normal Bearing No load Faulty Bearing Half load Faulty Bearing Full load Faulty Bearing

9. Bearing Specification Deep Groove Ball 6205 Bearing with Inside Diameter 25 mm, Outside Diameter 52 mm, Width 15 mm

10. Bearing condition Normal Bearing used in Experiment Faulty Bearing used in Experiment

11. NO Load Normal

12. Half Load Normal

13. Full Load Normal

14. No Load Faulty Bearing

15. Half Load Faulty Bearing

16. Full Load Faulty Bearing

17. Graph with Acceleration Data

18. No load Normal Bearing

19. Half load Normal Bearing

20. Full load Normal Bearing

21. No load Faulty Bearing

22. Half load Faulty Bearing

23. Experiment validation Total Sample collected - 17720 sample Time domain feature extraction method used Extracted 9 features for each axis data for each case of load and fault Dimensional features - Maximum , Minimum, Mean, Standard Deviation ,RMS, Skewness , Kurtosis Dimensionless features - Crest Factor, Form Factor

24. Experiment validation Reshape the dataset by 100 samples hence total dataset samples is 525 The reshaped dataset of 525 samples is divided into 80 % for training dataset and 20 % for testing dataset. Apply various Machine learning algorithm to Reshaped dataset

25. Algorithm used for Experiment Validation SVM - Support Vector Machine NBC- Naive Bayes classifiers K-NN- K - Nearest Neighbor DT- Decision Tree RF- Random Forest Python is used as a diagnostic tool

26. PROPOSED RESEARCH WORK Research work mainly focuses on design system using deep learning method to detect and predict the fault in rotating part of Induction Motor . Integrates a combination of Classification and Anomaly Detection algorithms Display the result about fault

27. Result of Each Algorithm confusion matix of SVM algorithm with kernel linear and random state=0

28. confusion matix of SVM algorithm with kernel polynomial and degree=4

29. confusion matix of SVM algorithm with kernel RBF and Random state =0

30. confusion matix of SVM algorithm with kernel Sigmoid and Random state =0

31. SVM algorithm results with accuracy of fault prediction Sr.No Algorithm Kernel Randome state / Degree Accuracy (%) 1 SVM Linear 26 2 SVM Poly 4 11 3 SVM RBF 10 4 SVM Sigmoid 7

32. Result of Naïve Bayes algorithm confusion matix of Naïve Bayes algorithm

33. Result of K-NN algorithm algorithm confusion matix of KNN algorithm

34. Result of DT algorithm confusion matrix of DT algorithm

35. Result of Random Forest algorithm confusion matrix of Random Forest algorithm

36. Overall Result Sr.No Algorithm Kernel Randome state / Degree Accuracy (%) 1 SVM Linear 26 2 SVM Poly 4 11 3 SVM RBF 10 4 SVM Sigmoid 7 Naïve Bayes 76 K-NN 11 Decision Tree 86.66 Random Forest 92.38

37. Conclusion Signal processing technique used for vibration analysis in Frequency domain or in time domain using various tool like Kurtogram , Cepstrum , FFT, Wavelets, Envelope Analysis. But input to this technique required sensor data and data acquisition hardware. Machine Learning can be a Supervised or Unsupervised and uses data from various data sources like Microsoft Research Open data, Google's Dataset Search Engine, Kaggle Datasets,UCIMachineLearning Repository,AWS If size of data is small and it is clearly labelled data then Supervised Learning is used. If size of data large then Unsupervised Learning method used to give better performance and results .If a huge data set easily available, then deep learning techniques is used for better accuracy.

38. REFERENCES [ 1] Filippetti F, Franceschini G, Tassoni C, Vas P AI techniques in induction machines diagnosis including the speed ripple effect. IEEE Trans Ind Appl 34:98–108 (1998). [2]Ludwig, Melanie, et al. "Measurement, Prediction, and Control of Individual Heart Rate Responses to Exercise—Basics and Options for Wearable Devices." Frontiers in physiology 9 (2018). [3] Tandon , N., and A. Choudhury. "A review of vibration methods for the detection of defects in rolling element bearings." Tribology international 32.8 (2011): 469-480. [4] Antoni, J. Fast computation of the kurtogram for the detection of transient faults. Mechanical Systems and Signal Processing, 21(1),108–124. (2007). [5]R. B. Randall and J. Antoni, “Rolling element bearing diagnostics-A tutorial,” Mech. Syst. Signal Process., vol. 25, no. 2, pp. 485–520, 2011. [6] W. He, Z. N. Jiang, and K. Feng, “Bearing fault detection based on optimal wavelet filter and sparse code shrinkage,” Meas. J. Int. Meas. Confed., vol. 42, no. 7, pp. 1092–1102, 2009. [7]I.S. Bozchalooi and M. Liang, “A smoothness indexguided approach to wavelet parameter selection in signal de-noising and fault detection,” J. Sound Vib ., vol. 308, pp. 246–267, 2007 . [8] A. Djebala , N. Ouelaa , and N. Hamzaoui , “Detection of rolling bearing defects using discrete wavelet analysis,” Meccanica , vol. 43, no. 3, pp. 339–348,2008. [9] P. K. Kankar , S. C. Sharma, and S. P. Harsha,“Rolling element bearing fault diagnosis using wavelettransform ,” Neurocomputing , vol. 74, no. 10, pp.1638–1645, 2011. [10] R. Dwyer, “Detection of non-Gaussian signals by frequency domain Kurtosis estimation,” ICASSP ’83. IEEE Int. Conf. Acoust . Speech, Signal Process., vol. 8, pp. 607–610, 1983

39. REFERENCES [ 11] D. Wang, P. W. Tse , and K. L. Tsui , “An enhanced Kurtogram method for fault diagnosis of rolling element bearings,” Mech. Syst. Signal Process., vol. 35, no. 1–2, pp. 176–199, Feb. 2013 [12] B.P. Bogert , M.J.R. Healy, J.W. Tukey, “The quefrency analysis of time series for echoes: cepstrum , pseudo- autocovariance , cross- cepstrum and saphe cracking,” in: Proceedings of Symposium on Time Series Analysis (1963). pp. 209–243 . [13] Choon -Su Park, Young- ChulChoi , Yang- HannKim , “Early fault detection in automotive ball bearings using the minimum variance cepstrum ” Mechanical Systems and Signal Processing 38 (2013) 534–548 [14] I. Bediaga , X. Mendizabal , A. Arnaiz , and J. Munoa , “ Ballbearing damage detection using traditional signal processingalgorithms ,”IEEE Instrumentation and Measurement Magazine,vol.16,no.2,pp.20–25,2013   [ 15] Mallat , S. A Wavelet Tour of Signal Processing; Academic Press: Burlington, VT, USA, 2009   [ 16] Peng, Z.K.; Chu, F.L. Application of the wavelet Transform in machine condition monitoring and fault diagnostics: A review with bibliography. Mech. Syst. Signal Process. 2004, 18, 199–221.  

40. REFERENCES [17] Srinivas, H.K.; Srinivasan, K.S.; Umesh , K.N. Role of an Artificial Neural Network and a Wavelet Transform for Condition Monitoring of the Combined Faults of Unbalance and Cracked Rotors. Int. J. Acoust . Vib . 2010, 15, 121–127.   [ 18] Hu, Q.; He, Z.; Zhang, Z. Fault diagnosis of rotating machinery based on improved wavelet package transform and SVMs ensemble. Mech. Syst. Signal Process. 2007, 21, 688–705   [ 19] Shen, C.; Wang, D.; Kong, F.; Tse , P.W. Fault diagnosis of rotating machinery based on the statistical parameters of wavelet packet paving and a generic support vector regressive classifier. Meas. J. Int. Meas. Confed. 2013, 46, 1551–1564.   [ 20] M Lokesha and K. Ramachandran, “Fault diagnosis in gear using wavelet envelope power spectrum,” Int. J. Eng. Sci. Technol., vol. 3, no. 8, Aug. 2012 . [21]Piotr Lipinski ,Edyta Brzychczy and Radoslaw Zimroz“Decision Tree-Based Classification for Planetary Gearboxes’ Condition Monitoring with the Use of Vibration Data in Multidimensional Symptom Space ,” Sensors 2020, 20, 5979.   [ 22]S. Ray, "A Quick Review of Machine Learning Algorithms," 2019 International Conference on Machine Learning, Big Data, Cloud and Parallel Computing ( COMITCon ), 2019,pp.35-39   [ 23] Cristianini,N ., and Shawe -Taylor, J., An Introduction to Support Vector Machines and Other Kernel-Based Learning Methods, Cambridge, UK: Cambridge University Press, 2000. [24] Shrinathan Esakimuthu Pandarakone1 Yukio Mizuno1and Hisahide Nakamura “A Comparative Study between Machine Learning Algorithm and Artificial Intelligence Neural Networking Detecting Minor Bearing Fault of Induction Motors”. Energies2019,12, 2105.  

41. REFERENCES [25]W. Zhou and B. Lu, “Incipient bearing fault detection via motor stator current noise cancellation using wiener filter,”IEEE Trans on Industry., vol. 45, no. 4, 2009, pp. 1309–131 [26] S. Dong, D. Sun, B. Tang, Z. Gao, W. Yu, and M. Xia, “A Fault Diagnosis Method for Rotating Machinery Based on PCA and Morlet Kernel SVM,” Mathematical Problems in Engineering, vol. 2014, 2014, pp. 1-8 .   [ 27] C.T. Yiakopoulos , K.C. Gryllias , I.A. Antoniadis “Rolling element bearing fault detection in industrial environments based on a K-means clustering approach “0957-4157 2010 Elsevier Ltd   [ 28] Yann LeCun and Yoshua Bengio . 1995. Convolutional networks for images, speech, and time series.Handbook ofBrain Theory and Neural Networks3361, 10 (1995), 255–257   [ 29] LeCun Y, Bengio Y, Hinton G (2015) Deep learning. Nature 521:1–10 [30] Shaveta Dargan, Munish Kumar,Maruthi Rohit Ayyagari,Gulshan Kumar “A Survey of Deep Learning and Its Applications: A New Paradigm to Machine Learning” Archives of Computational Methods in Engineering, Springer June 2019.   [ 31] Shen Zhang, Shibo Zhang,Bingnan Wang and Thomas G. Habetler “Deep Learning Algorithms for Bearing FaultDiagnostics – A Comprehensive Review” arXiv:1901.08247v3 [ cs.LG ] 6 Feb 2020

42.     [32] Si-Yu Shao,Wen -Jun Sun,Ru-Qiang Yan, Peng Wang, Robert X Gao “A Deep Learning Approach for Fault Diagnosis of Induction Motors in Manufacturing” Chin. J. Mech. Eng. (2017) 30:1347–1356 Springer October 2017   [33] Honglak Lee, Peter Pham, and Andrew Y Ng. Unsupervised feature learning for audio classification using convolutional deep belief networks. Pages 1-9, 2009a. [34] P.W.Tse , D. P. Atherton “Prediction of Machine Deterioration Using Vibration Based Fault Trends and Recurrent Neural Networks” Journal of Vibration and Acoustics Copyright © 1999 by ASME JULY 1999, Vol. 121 / 355

43. THANK YOU