Title :
Application of the variable architecture BP neural network in fault diagnosis of control system
Author_Institution :
Sch. of Electron. Eng., Yunnan Polytech. Univ., Kunming, China
Abstract :
In this paper a variable architecture BP neural network model is proposed. This new model possesses advantages over the previous one in operation speed, learning ability and fault tolerance to input information. It can be used to promote the recognizing rate of fault pattern and decrease the neural networks learning time in fault diagnosis of control system. The experimental results show that the model is feasible and effective
Keywords :
backpropagation; control systems; fault diagnosis; feedforward neural nets; neural net architecture; pattern recognition; BP neural network; backpropagation; control system; fault diagnosis; fault tolerance; learning; multilayer neural networks; pattern recognition; variable architecture; Artificial neural networks; Control system synthesis; Control systems; Electric variables control; Fault diagnosis; Intelligent networks; Multi-layer neural network; Neural networks; Neurons; Pattern recognition;
Conference_Titel :
Industrial Technology, 1996. (ICIT '96), Proceedings of The IEEE International Conference on
Conference_Location :
Shanghai
Print_ISBN :
0-7803-3104-4
DOI :
10.1109/ICIT.1996.601693
