Hidden Markov Model-Based Detection and Classification of Foreign Objects in Heat-Exchanger Tubes



Published Oct 3, 2016
Portia Banerjee Lalita Udpa Satish Udpa


In recent years, there has been significant interest in prognosis and health management of heat exchanger tubes in steam generators (SG) using eddy current (EC) non-destructive evaluation (NDE) techniques. One of the recent challenges encountered in SG tube inspection is the presence of foreign objects lodged outside the tubes. Extreme vibrations cause these loose parts to rub against the tube wall and form wears on their outer surfaces which can be dangerous in the high pressure environment. Hence, there is a strong need for reliable automated signal analysis systems for early detection of foreign objects and prevention of harmful radioactive leaks at nuclear facilities. In this paper, a hidden Markov model (HMM) based classifier is proposed which can estimate the material of the foreign object from EC inspection signal. Unknown loose part material interferes with EC analysis results and lead to errors in signal processing parameters which in turn can degrade performance and reliability of automated detection systems. The proposed algorithm implements a continuous HMM classifier by using magnitude and phase based measurements obtained from the foreign object. Results of applying the algorithm on experimental data from SG tube inspection is presented, demonstrating its benefits in increasing the robustness and performance of automated signal analysis systems in detecting loose parts.

How to Cite

Banerjee, P., Udpa, L., & Udpa, S. (2016). Hidden Markov Model-Based Detection and Classification of Foreign Objects in Heat-Exchanger Tubes. Annual Conference of the PHM Society, 8(1). https://doi.org/10.36001/phmconf.2016.v8i1.2557
Abstract 248 | PDF Downloads 104



classification, eddy current, Health Monitoring System, non-destructive testing, Hidden Markov Model, naive Bayes

Addin, O., Sapuan, S., Mahdi, E., & Othman, M. (2007). A na¨ıve-bayes classifier for damage detection in engineering materials. Materials & design, 28(8), 2379–2386.
Banerjee, P., Safdarnejad, S., Udpa, L., & Udpa, S. (2016). Ensemble of classifiers for confidence-rated classification of nde signal. In 42nd annual review of progress in quantitative nondestructive evaluation: Incorporating the 6th european-american workshop on reliability of nde (Vol. 1706, p. 180001).
Bilmes, J. A., et al. (1998). A gentle tutorial of the em algorithm and its application to parameter estimation for gaussian mixture and hidden markov models. International Computer Science Institute, 4(510), 126.
Boyles, R. A. (1983). On the convergence of the em algorithm. Journal of the Royal Statistical Society. Series B (Methodological), 47–50.
De Mesquita, R. N., Ting, D. K., Cabral, E. L., & Upadhyaya, B. R. (2004). Classification of steam generator tube defects for real-time applications using eddy current test data and self-organizing maps. Real-Time Systems, 27(1), 49–70.
Eren, L., & Devaney, M. J. (2004). Bearing damage detection via wavelet packet decomposition of the stator current. Instrumentation and Measurement, IEEE Transactions on, 53(2), 431–436.
Féron, O., & Mohammad-Djafari, A. (2004). A hidden markov model for bayesian data fusion of multivariate signals. arXiv preprint physics/0403149.
Gauvain, J.-L., & Lee, C.-H. (1994). Maximum a posteriori estimation for multivariate gaussian mixture observations of markov chains. IEEE transactions on speech and audio processing, 2(2), 291–298.
Grimberg, R., Udpa, L., Bruma, A., Steigmann, R., Savin, A., & Udpa, S. S. (2011). Eddy current examination of steam generator tubes from phwr power plants using rotating magnetic field transducer. International Journal of Microstructure and Materials Properties, 6(3-4), 307–322.
Hartigan, J. A., & Wong, M. A. (1979). Algorithm as 136: A k-means clustering algorithm. Journal of the Royal Statistical Society. Series C (Applied Statistics), 28(1), 100–108.
Haussler, D. K. D., & Eeckman, M. G. R. F. H. (1996). A generalized hidden markov model for the recognition of human genes in dna. In Proc. int. conf. on intelligent systems for molecular biology, st. louis (pp. 134–142).
Jarmulak, J. (1997). A method of representing and comparing eddy current lissajous patterns. In Review of progress in quantitative nondestructive evaluation (pp. 303–308). Springer.
Joo, K. M., & Shina, K. S. (2012). A development of techniques detecting foreign objects in the secondary side of steam generator. Transactions of the Korean Nuclear Society Autumn Meeting.
Jordan, M. I., & Xu, L. (1995). Convergence results for the em approach to mixtures of experts architectures. Neural networks, 8(9), 1409–1431.
Lee, J.W., Kirikera, G. R., Kang, I., Schulz, M. J., & Shanov, V. N. (2006). Structural health monitoring using continuous sensors and neural network analysis. Smart Materials and Structures, 15(5), 1266.
Mayo, C. W., & Shugars, H. G. (1988). Loose part monitoring system improvements. Progress in Nuclear Energy, 21, 505–513.
Mouret, M., Solnon, C., & Wolf, C. (2009). Classification of images based on hidden markov models. In Contentbased multimedia indexing, 2009. cbmi’09. seventh international workshop on (pp. 169–174).
Purushotham, V., Narayanan, S., & Prasad, S. A. (2005). Multi-fault diagnosis of rolling bearing elements using wavelet analysis and hidden markov model based fault recognition. Ndt & E International, 38(8), 654–664.
Qiu, X., Zhang, P., Wei, J., Cui, X., Wei, C., & Liu, L. (2013). Defect classification by pulsed eddy current technique in con-casting slabs based on spectrum analysis and wavelet decomposition. Sensors and Actuators A: Physical, 203, 272–281.
Rabiner, L. R. (1989). A tutorial on hidden markov models and selected applications in speech recognition. Proceedings of the IEEE, 77(2), 257–286.
Rish, I. (2001). An empirical study of the naive bayes classifier. In Ijcai 2001 workshop on empirical methods in artificial intelligence (Vol. 3, pp. 41–46).
Seymore, K., McCallum, A., & Rosenfeld, R. (1999). Learning hidden markov model structure for information extraction. In Aaai-99 workshop on machine learning for information extraction (pp. 37–42).
Sohn, H., Allen, D. W., Worden, K., & Farrar, C. R. (2005). Structural damage classification using extreme value statistics.
Udpa, L., Ramuhalli, P., Benson, J., & Udpa, S. (2004). Automated analysis of eddy current signals in steam generator tube inspection. Proceedings of the 16th WCNDT.
Xiang, P., Ramakrishnan, S., Cai, X., Ramuhalli, P., Polikar, R., Udpa, S., & Udpa, L. (2000). Automated analysis of rotating probe multi-frequency eddy current data from steam generator tubes. International Journal of Applied Electromagnetics and Mechanics, 12(3, 4), 151–164.
Zhou, W., Chakraborty, D., Kowali, N., Papandreou-Suppappola, A., Cochran, D., & Chattopadhyay, A. (2007). Damage classification for structural health monitoring using time-frequency feature extraction and continuous hidden markov models. In Signals, systems and computers, 2007. acssc 2007. conference record of the forty-first asilomar conference on (pp. 848–852).
Technical Research Papers