Back to home
 
Application of blind source separation to fault diagnosis of nuclear power equipment
 
JIANG Yingying1, XIA Hong2,and ZHU Shaomin3
 
1. College of Nuclear Science and Technology, Harbin Engineering University, Harbin, China (Tel: 18845895857, E-mail: 18845895857@163.com)
2. College of Nuclear Science and Technology, Harbin Engineering University, Harbin, China (Tel: 18645148138, E-mail: xiahong@hrbeu.edu.cn)
3. College of Nuclear Science and Technology, Harbin Engineering University, Harbin, China (Tel: 18945050958, E-mail: zsmtrue@163.com)
 
Abstract:Abstract:Nuclear power, which has been developing for 86 years, has become a major clean energy for human. With the development of nuclear power equipment, the requirements for safety andreliability of nuclear power equipment keep increasing. To ensure the safety and stable status of a nuclear power plant, it is necessary to constantly monitorequipment operational condition, timely discover abnormal operation of nuclear powerequipment and correctly diagnose the fault of nuclear power equipment. Signal processing, feature extraction, pattern recognition, and decision making are included in Fault diagnosis. The premise and foundation of these four steps is signal processing. In this paper, blind source separation based on artificial bee colony algorithm is applied to the signal processing in order to overcome the complex environment of the nuclear power plant and difficulty in signal extraction. Initialization, update strategy and adjustment strategy of the traditional artificial bee colony algorithm is optimized, and then the convergence speed and stability of the artificial bee colony algorithm are improved. Finally, simulation experiment results show that the algorithmproposed in this paper has better stability, convergence speed and global ergodicity comparedwith the traditional artificial bee colony algorithm, which also means the algorithm is feasible.
Keyword:fault diagnosis; blind source separation; artificial bee colony
 
      Click here for viewing full text.