基于T-S逻辑的新型模糊神经网络模型.PDFVIP

6 3 () Vo.l 6 N o. 3 2006 9 JOURNAL OF NAN JING NORMAL UN IVERSITY ( ENGINEERING AND TECHNOLOGY) Sep, 2006 基于T - S逻辑的新型模糊神经网络模型 1 2 姜长元, 朱庆保 ( 1. , 313000; 2. , 210097) [] , . , , , Stone-W eirstrass. . [] , , T- S [] TP183 [ ] A [ ] 1672-1292( 2006) 03-005 1-05 A New Fuzzy NeuralNetworkM odelBased on T-S Logic 1 2 J IANG C hangy uan , ZHU Q ingbao ( 1. Fa ulty of S ien e, H uzhou Tea hers C ollege, H uzhou 313000, Ch ina; 2. S hool of Mathem ati s and Com puter S ien e, Nanjing Norma lU n iversity, N anjing 210097, Ch ina) Abstract: Fuzzy neural newt ork, om posed of the neural netw ork and fuzzy logi system, is good at self- learn ing and self-tun ing. A ording to the inner relationsh ip betw een the neural newt ork and fuzzy logi inferen e, a new fuzzy neural netw ork an be ob tained w ith effe tive algorithm and stru ture. It an be proved to approxmi ate at any degree of a ura y to any real on tinuous fun tion in a ompa t dom ain by using the famous S tone-W eirstrass law. The smi u- lation shows that them odel is effe tive and feasible. K ey words: fuzzy logi , fuzzy neural newt ork, T - S logi 0 引言 , , , . [ 1] Takagi-Sugeno , . KOSKO B[ 2] , . J Zhang[ 3] , . CHEN Y C[ 4] . , , . [ 5, 6] , , , , . 1 模型原理 5BP, 1, : 2005-11-20. : ( 1975-) , , . E-m ail: j y@ hut . zj. n : ( 1955-) , , . E-m ai:l zhuqingb ao@ njnu. edu. n