Prognosis of Failure Time in Gas Turbines regarding multi time scales
Subject Areas : ForecastingTahminre Ashouri Moghadam 1 , Rasool Noorossana 2 , Sadigh Raissi 3
1 - Industrial Engineering Department, South Tehran Branch, Islamic Azad University, Tehran, Iran
2 - Iran University of Science and Technology
Tehran,Iran
3 - Industrial Engineering Department, South Tehran Branch, Islamic Azad University, Tehran, Iran
Keywords: gas turbine, Prognosis approach, wear pattern, Failure time,
Abstract :
In predicting of failure time based on operating hours, after identifying the effective parameters on wear process, it’s of great importance to model an associated function between parameters and their effects on acceleration the failure. Gas turbines are too complicated to simple functions, proposed by manufacturer, cover all the possible scenarios with high confidence. In this article, all the parameters affected by time and the working cycle are combined by a relation with regarding a correction function that improve the accuracy of forecast to calculate equivalent working hours. The approach presented here, all the parameters, with a defined coefficient affect the failure process and lifetime reduction of the system and the various wear fractions within the exclusive temperature range are weighed by an appropriate factor. This method helps give a more realistic estimate of the residual life with more precision. The correction function will be updated consequently base on accurate of last output in compare with actual result. Looking at the data, we can observe that 90 percent of failures happen in the area between 7800 to 8100 hours of work, which indicate accuracy with regard to these values given for the parameters. Also result demonstrates that consideration of the last corrections in lifetime prediction is more effective than consideration of all the available data.
[1] Li, Y.G., Nilkitsaranout, P. (2009). Gas turbine performance prognostic for condition-based maintenance. Applied Energy 86, 2152– 2161.
[2] Brotherton T, Grabill P, Wroblewski D, Friend R, Sotomayer B, Berry J. A testbed for data fusion for engine diagnostics and prognostics. In: Aerospace conference proceedings, 2002 IEEE, vol. 6; 2002. p. 3029–42.
[3] Tsui, K.L., Chen, N, Zhou, Q., Hai, Y., Wang, W. (2015). Prognostics and Health Management: A Review on Data Driven Approaches. Mathematical Problems in Engineering. ID 793161, 17 P.
[4] Xi, Z., Jing, R., Wang, P., & Hu, C. (2014). A copula-based sampling method for data-driven prognostics. Reliability Engineering & System Safety, 132, 72-82.
[5] Kalbfleisch, J. D., & Prentice, R. L. (2011). The statistical analysis of failure time data (Vol. 360). John Wiley & Sons.
[6] Song, S., Coit, D.W., Feng, Q., & Peng, H. (2014). Reliability analysis for multi-component systems subject to multiple dependent competing failure processes. Reliability, IEEE Transactions on, 63, 331–345.
[7] Kleinbaum, D., & Klein, M. (2011). Survival analysis: A selflearning text, 2005. New York Springer-Verlag.
[8] Zhang, Q., Hua, C., & Xu, G. (2014). A mixture Weibull proportional hazard model for mechanical system failure prediction utilizing lifetime and monitoring data. Mechanical
Systems and Signal Processing, 43, 103–112.
[9] Liu, X., Li, J., Al-Khalifa, K. N., Hamouda, A. S., Coit, D. W., & Elsayed, E. A. (2013). Condition-based maintenance for continuously monitored degrading systems with multiple failure modes. IIE Transactions, 45, 422–435.
[10] Ragab, A., Yacout, S., Ouali, MS. et al. (2019). Prognostics of multiple failure modes in rotating machinery using a pattern-based classifier and cumulative incidence functions. J Intell Manuf 30, 255– 274.
[11] Noorossana. R., Sabri-Laghaie. K. (2015). System reliability with multiple failure modes and time scales. Quality and reliability engineering international, 1-18.
[12] Singpurwalla, N.D. and Wilson. S. P. (1998). Failure models indexed by two scales .Advances in Applied Probability. 30(4): p. 1058- 1072.