Your browser does not support JavaScript!
http://iet.metastore.ingenta.com
1887

access icon free Combining and comparing various machine-learning algorithms to improve dissolved gas analysis interpretation

© The Institution of Engineering and Technology
Received 12/01/2018, Accepted 19/05/2018, Revised 26/04/2018, Published 24/05/2018

Since the discovery of dissolved gas analysis (DGA), it is considered as a leading technique for the diagnosis of liquid insulated power equipment. However, accurate analysis results can only be achieved if the measured gases closely reflect the actual equipment condition to enable an appropriate interpretation of these gases. In general, conventional techniques such as the ratio method, key gases, and Duval triangle combined or not with artificial intelligence techniques such as machine-learning algorithms are used for DGA interpretation. Here, four well-known machine-learning algorithms are compared in terms of DGA fault classification – Bayes network, multilayer perceptron, k-nearest neighbour, and J48 decision tree. Moreover, the effect of applying ensemble methods such as boosting through the Adaboost algorithm and bootstrap aggregation (bagging) is analysed, and the performances of these algorithms are evaluated. The data for developing classification models was transformed into three forms, other than the raw data. The obtained results clearly presented the efficiency and stability of some algorithms such as the J48 tree and Bayes networks for DGA fault classification, in particular, when the data is appropriately pre-processed. Moreover, the performance of these algorithms was found to consistently improve by integrating the concepts of multiple models or ensemble methods.

Inspec keywords: Bayes methods; power engineering computing; pattern classification; decision trees; transformer oil; fault diagnosis; chemical engineering computing; multilayer perceptrons; nearest neighbour methods

Other keywords: artificial intelligence techniques; DGA interpretation; bootstrap aggregation; classification models; Bayes network; ratio method; k-nearest neighbour; Adaboost algorithm; dissolved gas analysis interpretation; multilayer perceptron; J48 decision tree; machine-learning algorithms; key gases; DGA fault classification; Duval triangle

Subjects: Chemical engineering computing; Organic insulation; Data handling techniques; Power engineering computing; Other topics in statistics; Neural computing techniques; Combinatorial mathematics; Combinatorial mathematics; Other topics in statistics; Knowledge engineering techniques

References

    1. 1)
      • 21. Malik, H., Khatri, A., Dohare, R.: ‘Probabilistic neural network based incipient fault identification using DGA dataset’, Procedia Comput. Sci., 2015, 58, pp. 655–672.
    2. 2)
      • 30. Boudraa, S.: ‘Dissolved gas in transformer oil analysis using artificial intelligence’. Master Thesis, Batna University, Algeria, 2005.
    3. 3)
      • 38. LiorRokach: ‘Ensemble-based classifiers’, Artif. Intell. Rev., 2010, 33, pp. 1–39.
    4. 4)
      • 39. Ali, S., Tirumala, S.S., Sarrafzadeh, A.: ‘Ensemble learning methods for decision making: status and future prospects’. Int. Conf. Machine Learning and Cybernetics, Guangzhou, 2015, pp. 211–216.
    5. 5)
      • 3. Sun, H.-C., Huang, Y.-C., Huang, C.-M.: ‘A review of dissolved gas analysis in power transformers’, Energy Procedia, 2012, 14, pp. 1220–1225.
    6. 6)
      • 19. Nagpal, T., Brar, Y.S.: ‘Neural network based transformer incipient fault detection’. IEEE Int. Conf. Advances in Electrical Engineering, January 2014, pp. 1–5.
    7. 7)
      • 34. Graczyk, M., Lasota, T., Trawiński, B., et al: ‘Comparison of bagging, boosting and stacking ensembles applied to real estate appraisal, intelligent information and database systems’. 2nd Int. Conf., Hue City, Vietnam, March 2010.
    8. 8)
      • 17. Mirowski, P., LeCun, Y.: ‘Statistical machine learning and dissolved gas analysis: a review’, IEEE Trans. Power Deliv., 2012, 27, (4), pp. 1791–1799.
    9. 9)
      • 18. Salami, A., Pahlevani, P.: ‘Neural network approach for fault diagnosis of transformers’. Int. Conf. Condition Monitoring and Diagnosis, Beijing, China, April 2008, pp. 1346–1349.
    10. 10)
      • 16. Bakar, N., Abu-Siada, A., Islam, S.: ‘A review of dissolved gas analysis, measurement and interpretation techniques’, IEEE Electr. Insul. Mag., 2014, 30, (3), pp. 39–49.
    11. 11)
      • 7. IEC 60599: ‘Mineral-oil-impregnated electrical equipment service – guide to the interpretation of dissolved and free gas analysis’, 1999.
    12. 12)
      • 15. Duval, M., Lamarre, L.: ‘The Duval pentagon – a new complementary tool for the interpretation of dissolved gas analysis in transformers’, IEEE Electr. Insul. Mag, 2013, 30, (6), pp. 9–12.
    13. 13)
      • 8. Fofana, I., Sabau, J., Bussières, D., et al: ‘The mechanism of gassing in power transformers’. 16th IEEE Int. Conf. Dielectric Liquids, Poitiers, France, 30 June – 4 July 2008, pp. 1–4.
    14. 14)
      • 11. Krontiris, T.: ‘Fuzzy systems for condition assessment of equipment in electric power systems’. PhD Thesis, Darmstadt University of Technology, Germany, 2012.
    15. 15)
      • 22. Thang, K.F., Aggarwal, R.K.: ‘Analysis of power transformer dissolved gas data using the self-organizing map’, IEEE Trans. Power Del., 2003, 18, (4), pp. 1241–1248.
    16. 16)
      • 24. Souahlia, S., Bacha, K., Chaari, A.: ‘SVM-based decision for power transformers fault diagnosis using Rogers and Doemenburg ratios DGA’. 10th IEEE Int. Multi-Conf. Systems, Signals & Devices Hammamet, Tunisia, March 2013, pp. 1–6.
    17. 17)
      • 1. Fofana, I.: ‘50 years in the development of insulating liquids’, IEEE Electr. Insul. Mag., 2013, 29, (5), pp. 13–25.
    18. 18)
      • 32. Akbari, A., Setayeshmehr, A., Borsi, H., et al: ‘Intelligent agent-based system using dissolved gas analysis to detect incipient faults in power transformers’, IEEE Electr. Insul. Mag, 2010, 26, (6), pp. 27–40.
    19. 19)
      • 26. Islam, M.M, Lee, G., Hettiwatte, S.N.: ‘Incipient fault diagnosis in power transformers by clustering and adapted KNN’. Australasian Universities Power Engineering Conf., Brisbane, QLD, 2016, pp. 1–5.
    20. 20)
      • 31. Abidin, Z., Muhammad, S.H., Abdulrashid, K., et al: ‘Design of a fault diagnostic engine for power transformer using data mining’. Project Report. Faculty of Electrical Engineering, Skudai, Johor, 2007. Available at http://eprints.utm.my/5839/.
    21. 21)
      • 20. Souahlia, S., Bacha, S., Chaari, A.: ‘Power transformer fault diagnosis based on dissolved gas analysis by artificial neural network’. 1st IEEE Int. Conf. Renewable Energies and Vehicular Technology, Hammamet, Tunisia, March 2012, pp. 230–236 .
    22. 22)
      • 12. Duval, M.: ‘Dissolved gas analysis: It can save your transformer’, IEEE Electr. Insul. Mag., 1989, 5, (6), pp. 22–27.
    23. 23)
      • 28. Duval, M., De Pablo, A.: ‘Interpretation of gas-in-oil analysis using new IEC publication 60599 and IEC TC 10 databases’, IEEE Electr. Insul. Mag., 2001, 17, pp. 31–41.
    24. 24)
      • 4. Huang, Y.-C.: ‘A new data mining approach to dissolved gas analysis of oil-insulated power apparatus’, IEEE Trans. Power Del., 2003, 18, (4), pp. 1257–1261.
    25. 25)
      • 37. Ben-Gal, I.: ‘Bayesian networks’, Encyclopedia of Statistics in Quality and Reliability (John Wiley & Sons, Hoboken, NJ, USA, 2007).
    26. 26)
      • 13. Church, J.O., Haupert, T.J., Jakob, F.: ‘Analyze incipient faults with dissolved gas Nomograph’, Electrical world, 1987, 201, pp. 40–44.
    27. 27)
      • 2. Ian, A., Gray, R.: ‘Practical experience gained from dissolved gas analysis at an aluminium smelter’. Doble Eskom Annual Int. Conf., Southern Africa, 2010.
    28. 28)
      • 27. Lakehal, A., Ghemari, Z., Saad, S.: ‘Transformer fault diagnosis using dissolved gas analysis technology and Bayesian networks’. 4th Int. Conf. Systems and Control, Sousse, 2015, pp. 194–198.
    29. 29)
      • 5. Ahuja, H., Bhuvaneswari, G., Balasubramanian, R., et al: ‘Soft computing – applied to transformer incipient fault diagnosis’. IEEE 4th Int. Conf. Computational Intelligence and Communication Networks, 2012, pp. 809–814.
    30. 30)
      • 10. Wang, Z.: ‘Artificial intelligence applications in the diagnosis of power transformer incipient faults’. PhD Thesis, Virginia Polytechnic Institute and State University, 2000.
    31. 31)
      • 25. Zhang, M., Li, K., Tian, H.X.: ‘Multiple SVMs modelling method for fault diagnosis of power transformers’, Przeglad Elektrotechniczny, 2012, 88, (7), pp. 232–234.
    32. 32)
      • 35. Quinlan, J.R.: ‘C4.5: programs for machine learning’ (Morgan Kaufmann Publishers, San Mateo, 1993).
    33. 33)
      • 6. ‘IEEE – institute of electrical and electronics engineer – IEEE guide for the interpretation of gases generated in oil immersed transformer’, IEEE Std C57.104, 2008.
    34. 34)
      • 36. Witten, I., Frank, E.: ‘Data mining: practical machine learning tools and techniques with java implementations’ (Morgan Kaufmann Press, San Francisco, CA, USA, 2005, 2nd edn.).
    35. 35)
      • 9. Tang, W.H., Wu, Q.H.: ‘Condition monitoring and assessment of power transformers using computational intelligence’ (Springer-Verlag London Limited, London, UK, 2011, 1st edn.).
    36. 36)
      • 29. Tomsovic, K., Amar, A.: ‘On refining equipment condition monitoring using fuzzy sets and artificial neural nets’, Int. J. Eng. Intell. Syst., 1997, 5, (1), pp. 43–50.
    37. 37)
      • 33. Wu, X.: ‘Top 10 algorithms in data mining’, Knowl. Inf. Syst., 2008, 14, pp. 1–37.
    38. 38)
      • 23. Singh, J., Kaur, K., Kumari, P., et al: ‘Condition assessment of power transformer using dissolved gas analysis’, Int. J. Adv. Res. Electr., Electron. Instrum. Eng., 2016, 5, (4), pp. 1–6.
    39. 39)
      • 14. Singh, S., Bandyopadhyay, M.N.: ‘Duval triangle-A noble technique for DGA in power transformers’, Int. J. Electr. Power Eng., 2010, 4, (3), pp. 193–197.
http://iet.metastore.ingenta.com/content/journals/10.1049/iet-gtd.2018.0059
Loading

Related content

content/journals/10.1049/iet-gtd.2018.0059
pub_keyword,iet_inspecKeyword,pub_concept
6
6
Loading
Print this page
This is a required field
Please enter a valid email address