基于精确罚函数的一类广义非线性神经网络模型1).PDFVIP

DOI :10.16383/j.aas.2003.05.016 29 　5 　　　　 Vol.29, No.5 2003 9 ACTA AUT OMAT ICA SINICA Sept., 2003 　 1) 1 1 2 孟志青 　　胡奇英 　　杨晓琪 1( 　 　710071) 2 (　) (E-mail:mengzhiqing @ tu.edu.cn) 　　2 , , , , . . 　, , , , 　O122.2 A General Model of Non-Linear Neural Networks Based on Exact Penalty Function 1 1 2 M ENG Zhi-Qing 　HU Qi-Ying 　YANG Xiao-Qi 1(S chool of Econom ics and M anagement, X idian University , X i ′an 　710071) 2 (D ep artmen t of Ap p lied Mathematics, The Hong K ong Poly technic University , Hong K ong ) (E-mail:mengzhiqing @ tu.edu.cn) Abstract　A double non-linear penalty function is defined for the non-linear optimality prob- lems (NP )and the e act penalty theorem is e acted under some conditions.A new general model of non-linear neural networks is introduced and the relationship between the equilibri- um points and the energy function is showed.Under the given condition, the equilibrium point of the neural networks converges to a solution of NP.This model plays an important part in many optimal problems. Key words 　Neural networks, non-linear penalty function, optimal solution, equilibrium point, stable point 1) Supported by National Natural Science Foundation of P.R.China 　2001-04-11　　　2001-11-15 Received April 11, 2001;in invised form Novem ber 15, 2001 756 　　　　　　　　 29 1