The Proceedings of the International Conference on Power Engineering (ICOPE)
Online ISSN : 2424-2942
2003.3
Conference information
B303 THE MONITORING MODEL AND ITS EXPERIMENTAL STUDY OF FAN PERFORMANCE BASED ON HYBRID NEURAL NETWORK
Lian suo ANHai yan HUSong ling WANGJun hu HOUXiong jun YUAN
Author information
CONFERENCE PROCEEDINGS FREE ACCESS

Pages _3-89_-_3-91_

Details
Abstract
Aiming at the centrifugal fan which is widely used in power plants, in this paper using the strong nonlinear approaching capacity of RBFNN model and good generalizing ability of mathematic model, a monitoring model of fan performance based on hybrid NN is presented, which combines RBFNN model and partial least squares (PLS) model. The performance curves of 4-73No.8D are approached by experimental data and study results show that the model has higher precision than normal RBFNN model. In the last part of the paper, the application of the model under different conditions in studied.
Content from these authors
© 2003 The Japan Society of Mechanical Engineers
Previous article Next article
feedback
Top