A Unified Approach to Implement Nonlinear Controllers Based on Neurofuzzy Networks  

Oleh:

K. Yeung, W. ; Chan C.W. ; Cheung K.C.

Jenis: Article from Article
Dalam koleksi: Final Program and Book of Abstracts: The 4th Asian Control Conference, September 25-27, 2002 (Sep. 2002), page 297-302.
Topik: Neurofuzzy networks; Nonlinear Controllers; GMV; Integrating control law
Fulltext: AC021093.PDF (127.83KB)

Isi artikel

Nonlinear controllers implemented by neurofuzzy networks are discussed in this paper. The structure of the controllers is derived using the same approach as that in deriving the generalized minimum variance (GMV) and the integrating control laws. The main advantage of using neurofuzzy networks to implement the nonlinear controllers is that they can be trained using a simplified recursive least squares method derived from the local change property of neurofuzzy networks. The proposed training algorithm reduces drastically the computing time and improves the tracking ability of the training algorithm. From the simulation example, it is shown that the neurofuzzy controllers derived from the integrating control law can eliminate offsets in the system, but not that derived from the GMV control law.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali