An accident diagnosis algorithm for HTR-PM based on deep learning methods

Abstract The Chinese High Temperature Reactor Pebble-bed Module (HTR-PM) is the world first commercial nuclear power plant (NPP) with the characteristics of fourth generation. Accident diagnosis tasks for HTR-PM are directly associated with safe and efficient operation. Although different kinds of accident diagnosis methods have been studies on conventional NPPs, the research of accident diagnosis of HTR-PM is relatively lacking according to the different characteristics and new applications of HTR-PM. In this article, a new algorithm for HTR-PM accident diagnosis based on deep learning methods is proposed. By using the preprocessing, classification network and postprocessing techniques, the proposed algorithm could avoid over-reliance on the previous experiences and make fully use of the signals, also it can use only few number of training signal sequences to get high accuracy results, which is a significant advantage compared with traditional algorithms using deep learning methods. This algorithm is tested using the signals produced by the engineering simulator of HTR-PM for normal state and several accidents, including loss of feed water, large break depressurized loss of forced cooling, small break depressurized loss of forced cooling and inadvertent withdrawal of a single control rod. The results show the feasibility and effectiveness of the algorithm for HTR-PM accident diagnosis, and also the potentiality to use in other NPPs accident diagnosis tasks.

[1]  Eric B. Bartlett,et al.  Nuclear power plant status diagnostics using an artificial neural network , 1992 .

[2]  Tao Zhang,et al.  Bearing fault diagnosis method based on stacked autoencoder and softmax regression , 2015, 2015 34th Chinese Control Conference (CCC).

[3]  A. Srivastava,et al.  Diagnostic system for identification of accident scenarios in nuclear power plants using artificial neural networks , 2009, Reliab. Eng. Syst. Saf..

[4]  Li Deng,et al.  Ensemble deep learning for speech recognition , 2014, INTERSPEECH.

[5]  J. K. Vaurio.  Safety-related decision making at a nuclear power plant , 1998 .

[6]  Man Gyun Na,et al.  Prediction of major transient scenarios for severe accidents of nuclear power plants , 2004 .

[7]  Demis Hassabis,et al.  Mastering the game of Go with deep neural networks and tree search , 2016, Nature.

[8]  Wondea Jung,et al.  A taxonomy of performance influencing factors for human reliability analysis of emergency tasks , 2003 .

[9]  Zuoyi Zhang,et al.  Economic potential of modular reactor nuclear power plants based on the Chinese HTR-PM project , 2007 .

[10]  I. Guyon,et al.  Handwritten digit recognition: applications of neural network chips and automatic learning , 1989, IEEE Communications Magazine.

[11]  Feng Jia,et al.  An Intelligent Fault Diagnosis Method Using Unsupervised Feature Learning Towards Mechanical Big Data , 2016, IEEE Transactions on Industrial Electronics.

[12]  U. S. Doe.  A Technology Roadmap for Generation IV Nuclear Energy Systems , 2002 .

[13]  Kee-Choon Kwon,et al.  Accident identification in nuclear power plants using hidden Markov models , 1999 .

[14]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[15]  Tom Kontogiannis,et al.  Stress and operator decision making in coping with emergencies , 1996, Int. J. Hum. Comput. Stud..

[16]  Jeffrey L. Elman,et al.  Finding Structure in Time , 1990, Cogn. Sci..

[17]  Hong Wang,et al.  The Shandong Shidao Bay 200 MW e High-Temperature Gas-Cooled Reactor Pebble-Bed Module (HTR-PM) Demonstration Power Plant: An Engineering and Technological Innovation , 2016 .

[18]  Yuanle Ma,et al.  The engineering simulation system for HTR-PM , 2014 .

[19]  Jaemin Yang,et al.  An accident diagnosis algorithm using long short-term memory , 2018 .

[20]  Qin Zhang,et al.  Pilot study of dynamic Bayesian networks approach for fault diagnostics and accident progression prediction in HTR-PM , 2015 .

[21]  Yujie Dong,et al.  Current status and technical description of Chinese 2 × 250 MWth HTR-PM demonstration plant , 2009 .

[22]  Z. Guo,et al.  Nuclear power plant performance study by using neural networks , 1991, Conference Record of the 1991 IEEE Nuclear Science Symposium and Medical Imaging Conference.

[23]  Mauro Vitor de Oliveira,et al.  HSI for monitoring the critical safety functions status tree of a NPP , 2013 .

[24]  David Meister,et al.  Cognitive behavior of nuclear reactor operators , 1995 .