Distribution power system outage diagnosis based on root cause analysis

Document Type : Article


School of Electrical and Computer Engineering, Department of Engineering, Shiraz University, Zand Avenue, Shiraz, P.O. Box 71348-51154, Iran.


This paper proposes data mining-based models to diagnose outage data in distribution power systems. In this work, outage data from a local distribution company is gathered and aligned with weather data. Then, a subset of features is selected to reduce the processing time and simplifying purposes. To increase the fairness of final models and to account for differences in misclassification cost, using a customized cost matrix is proposed. Two decision tree-based modeling algorithms are trained and tested. Results show the ability of the established models to diagnose the root cause of an outage fairly well. In addition, an ensemble of the decision tree-based models is built, which outperforms the other two models in almost all cases. Finally, applications of such models in decreasing outage duration and improving the reliability of the power distribution network are discussed.


1.Zheng, K., Chen, Q., Wang, Y., et al. A novel  combined data-driven approach for electricity theft  detection", IEEE Trans. Ind. Informatics, 15(3), pp.  1809-1819 (2018). 
2. Jokar, P., Arianpoo, N., and Leung, V.C.M. Electricity  theft detection in AMI using customers consumption  patterns", IEEE Trans. Smart Grid, 7(1),  pp. 216-226 (2016). 
3. Singh, S.K., Bose, R., and Joshi, A. PCA based  electricity theft detection in advanced metering infrastructure",  7th Int. Conf. Pow. Syst. (ICPS), pp. 441-  445 (2017). 
4. Venkatesh, A., Cokkinides, G., and Meliopoulos,  A.P.S. 3D-visualization of power system data using  triangulation and subdivision techniques", Proc. 42nd  Annu. Hawaii Int. Conf. Syst. Sci. HICSS, pp. 1-8  (2009).  5. McMorran, A.W., Ault, G.W., and McDonald, J.R.  Solving data integration challenges for web-based geographical  power system data visualization using CIM",  IEEE Power Energy Soc. 2008 Gen. Meet. Convers.  Deliv. Electr. Energy 21st Century, PES (2008).  6. Bank, J.N., Omitaomu, O.A., Fernandez, S.J., et  al. Visualization and classi_cation of power system  frequency data streams", ICDM Work. 2009 - IEEE  Int. Conf. Data Min., pp. 650-655 (2009).  7. Brown, R.E., Electric Power Distribution Reliability,  2nd Edn., CRC press (2008).  8. Schroder, T. and Kuckshinrichs, W. Value of lost  load: An e_cient economic indicator for power supply  security? A literature review", Front. Energy Res., 3,  p. 55 (2015).  9. Eskandarpour, R. and Khodaei, A. Machine learning  based power grid outage prediction in response to  extreme events", IEEE Trans. Power Syst., 32(4), pp.  3315-3316 (2017).  10. Wanik, D.W., Anagnostou, E.N., Hartman, B.M., et  al. Storm outage modeling for an electric distribution  network in Northeastern USA", Nat. Hazards, 79(2),  pp. 1359-1384 (2015).  11. Doostan, M. and Chowdhury, B.H. Power distribution  system fault cause analysis by using association  rule mining", Electr. Power Syst. Res., 152, pp. 140-  147 (2017).  12. Cai, Y. and Chow, M.Y. Cause-e_ect modeling and  spatial-temporal simulation of power distribution fault  events", IEEE Trans. Power Syst., 26(2), pp. 794-801  (2011).  13. Chow, M. and Taylor, L.S. Analysis and prevention  of animal-caused faults in power distribution systems",  IEEE Trans. Power Deliv., 10(2), pp. 995-1001 (1995).  14. Warlyani, P., Jain, A., Thoke, A.S., et al. Fault  classi_cation and faulty section identi_cation in teed  transmission circuits using ANN", Int. J. Comput.  Electr. Eng., 3(6), pp. 807-811 (2012).  15. Xu, L., Chow, M.-Y., and Taylor, L.S. Power distribution  fault cause identi_cation with imbalanced  data using the data mining-based fuzzy classi_cation  E-Algorithm", IEEE Trans. Power Syst., 22(1), pp.  164-171 (2007).  16. Doostan, M. and Chowdhury, B.H. Power distribution  system equipment failure identi_cation using  machine learning algorithms", IEEE Pow. Energy Soc.  Gen. Meet., pp. 1-5 (2018)  17. Doostan, M., Sohrabi, R., and Chowdhury, B. A  data-driven approach for predicting vegetation-related  outages in power distribution systems", arXiv Prepr.  arXiv1807.06180, pp. 1-11 (2018).  18. Breiman, L. Random forests", Mach. Learn., 45(1),  pp. 5-32 (2001).  19. Quinlan, J.R. C5.0 data mining tool", RuleQuest  Research, 63, p. 64 (1997).  20. Quinlan, J.R. C 4.5: Programs for machine learning",  Morgan Kaufmann Ser. Mach. Learn. San Mateo, CA  Morgan Kaufmann (1993).  21. Hinton, G. and Van Camp, D. Keeping neural networks  simple by minimizing the description length of  the weights", Proc. 6th Ann. ACM Conf. on Computational  Learning Theory (1993).  22. Raspisaniye Pogodi Ltd, Reliable Prognosis. Weather  and Climate Change, Available Online at 2019:  https://rp5.ru/.  23. Liu, H. and Motoda, H. Feature selection for knowledge  discovery and data mining", 454, Springer Science  & Business Media (2012).  24. Guyon, I. and Elissee_, A. An introduction to variable  and feature selection", J. Mach. Learn. Res., 3(Mar),  pp. 1157-1182 (2003).