Application of Fuzzy Neural Network to Fault Diagnosis of Sensor with Mechanical Properties in Nuclear Power Plant

2013 ◽  
Vol 644 ◽  
pp. 68-71 ◽  
Author(s):  
Jin Yang Li ◽  
Hong Xia
Keyword(s):  
Neural Network ◽  
Mechanical Properties ◽  
Fault Diagnosis ◽  
Power Plant ◽  
Nuclear Power Plant ◽  
Nuclear Power ◽  
Fuzzy Neural Network ◽  
Back Propagation ◽  
Back Propagation Algorithm ◽  
Fuzzy Neural

In view of the sensor fault in nuclear power plant, it puts forward a method to fault diagnosis of sensor with mechanical properties based on fuzzy neural network. The method would be fuzzy logic control combined with neural network. It adjusted and corrected membership function parameters and network weights with back propagation algorithm. After the completion of fuzzy neural network training, it could get the credibility of sensor with mechanical properties real time. Taking pressurizer water-level sensor as the case, the simulation experiment results showed that the method is valid for the fault diagnosis of sensor with mechanical properties in nuclear power plant.

Study on Sensor with Mechanical Properties in Nuclear Power Plant with Application of BP Neural Network to Fault Tolerant Control

2013 ◽  
Vol 644 ◽  
pp. 56-59
Author(s):  
Jin Yang Li ◽  
Hong Xia ◽  
Shou Yu Cheng
Keyword(s):  
Neural Network ◽  
Mechanical Properties ◽  
Power Plant ◽  
Nuclear Power Plant ◽  
Bp Neural Network ◽  
Nuclear Power ◽  
Control Method ◽  
Fault Tolerant ◽  
Fault Tolerant Control ◽  
Level Sensor

All kinds of sensor with mechanical properties often can go wrong in nuclear power plant. In this kind of situation, it puts forward a kind of active fault tolerant control method based on the improved BP neural network. Firstly, the method will train sensor by BP neural network. Secondly, it will be established dynamic model bank in all kinds of running state. The system will be detected by using BP neural network real time. When the sensor goes wrong, it will be controled by reconstruction. Taking pressurizer water-level sensor as the case, a simulation experiment was performed on the nuclear power plant simulator. The results showed that the proposed method is valid for the fault tolerant control of sensor in nuclear power plant.

Fault diagnosis of nuclear power plant based on genetic-RBF neural network

2010 ◽  
Vol 39 (1/2/3) ◽  
pp. 159 ◽  
Author(s):  
Chun Ling Xie ◽  
Jen Yuan Chang ◽  
Xiao Cheng Shi ◽  
Jing Min Dai
Keyword(s):  
Neural Network ◽  
Fault Diagnosis ◽  
Power Plant ◽  
Nuclear Power Plant ◽  
Nuclear Power ◽  
Rbf Neural Network

A New Method of Complex Maneuvering Event Detection

2011 ◽  
Vol 63-64 ◽  
pp. 313-317
Author(s):  
Wei Chen ◽  
Qing Li ◽  
Jian Bin He ◽  
Tao Jin
Keyword(s):  
Neural Network ◽  
Real Time ◽  
Fuzzy Neural Network ◽  
Rough Set Theory ◽  
Back Propagation ◽  
Neural Cells ◽  
Back Propagation Algorithm ◽  
Fuzzy Neural ◽  
Rule Sets ◽  
Events Detection

In dealing with such problems as imprecision and poor real-time performance in complex maneuverable events detection. A method based on RFNN (Rough-Fuzzy Neural Network) is proposed. Firstly, the minimal rule sets from data samples are acquired by using the Rough Set Theory; secondly, these rules are used to construct the initial scalar values of neural cells in each layer and their relative parameters in the fuzzy neural network; lastly, parameters of the network are acquired by using BP(back propagation) algorithm. The experimental results indicate that RFNN take advantage of the sample data features effectively, reduce the number of rules, simplify network structure, improve the precision of detection and the performance of real-time.

Sign in / Sign up

Export Citation Format

Share Document