Ϣ
INFORMATION AND CONTROL
1999 28 3 Vol.28 No.3 1999



CMAC(Сģ)񾭼񾭿
ۻݺ
ժҪCMACǾֲѧϰ磬ṹ򵥣ٶȿ죬ͺӲʵ֣й㷺ӦǰCMACṹ㷨ԼڿеӦãָCMAC񾭼񾭿ƷչʵӦ⣮
ؼʣCMACѧϰ㷨񾭼㣬񾭿
ͼţTP11ױʶ룺A
CMAC NEURAL NETWORKS BASED NEURAL COMPUTATION
AND NEUROCONTROL
DUAN PeiyongSHAO Huihe
(Department of Automation, Shanghai Jiaotong University200030)
AbstractA CMAC is one kind of local-learning neural networks, it has promising applications in the future for its simple structure, fast convergence speed and easy implementation by software and hardware. Structures and algorithms of different CMAC are summarized, applications to control systems are also discussed. At the end of the paper, it is pointed out that neural computation and neurocontrol based on CMAC have several further research aspects and some problems to be overcome.
Key words CMAC, learning algorithm, neural computation, neurocontrol
1
CMACAlbus1Eccles2Сʱģ͵Ļϣ1975һģСԵѧϰṹСģ͹ؽڿ(Cerebellar Model Articulation Controller)ΪCMACкǿļAlbus CMACǻڱѯ(Table lookup)ʽľֲƽ磬һϢĴ洢ָģСԵ(Recall)򵥽ṹģͼ1ʾ


ͼ1Albus CMACģ
ռпܵxiɣCMAC罫ܵκ֪ͨMӳ䵽һܴ洢Aеc㣮ռǵcڴ洢AвصĵԪԽصԽ࣮֮ռԶǵcAûصΪ˼ʵʴ洢ռ䣬Ӵձ(Hashing-coding)洢ӳ䵽һСöʵֵĴ洢A䣮ôκCMAC뽫AcʵλãЩλ洢ȨӵõyMΪ֪cΪ֪ҰCMACʵֵ㷶(Generalization)(Dichotomy)ԭ޹
CMAC㷨ȽĲ֪ص㣬ٶȺܿ죬֤ЧڷԺƽ̬ͨӳʵֶ̬ģϵͳƵȣ类ڻ˿ƣ34ݡƣ5ݡѹ絶ܿƣ6ݡеϵͳƣ7ݡܾϵͳ8ݡźŴ9ݡȼϹϵͳƣ10ݡӡɫУ׼11ݵУ
ĿƣNN-based ControlΪ,񾭿ƣNeurocontrolNeurocontrolH. TolleErsu1992ʽܿƵһոµо򣬿ܳΪܿƵġ֮㡱1213ݣCMAC񾭿ƵоCMAC籾о;иCMACṹѧϰ㷨ĹϵǻCMACص㣬пƲԵо
2CMACԽ
񾭼񾭿ҪӦĺƽǶȿɷΪȫֱƽ;ֲƽ磮һͶȨϵӦɵռÿһκһӰ죬ƸΪȫֱƽ磬BPÿԣÿһȨеӶȫֱƽٶȺȱ㣮ȱڿ˵ǲ̵ģռĳֲֻȨӰƸΪֲƽ磬ÿݶԣֻȨҪеӶʹֲƽѧϰٶȿŵ㣬һڿ˵ҪģCMACBRBFԼĳЩģǾֲƽ磮
ϣ3ϵBPܹƽκһԺBPȫֱƽ磬ÿһѧϰҪµȨֵٶֲСϵͳʵʱҪCMACԽڣ1) CMACϢ洢ھֲĽṹϣڱ֤ƽܵǰ£ѧϰٶȿ죮2) CMACƽѧϰݳֵĴУCMACʱ䡢ϵͳеõ㷺Ӧã14ݣ3) CMACṹ򵥣Ӳʵ֣1516ݺʵ֣1718ݣ
3CMACĽṹ
3.1CMACṹ
Albus CMACṹ򵥣Ӳʵ磬Ӧýձ飮Ȩϵ洢ռCMACάӣӦϵһѣһĽǲӴձ루Hash-coding\[19\]о÷ѧϰʱײ⣬Ҫôѧϰٶ½ҪôѧϰɢWangָӴձ뷨CMACıƽ20ݣǴͳ۵㲻ͬоԱͼά⣨͵Ҳͬ⣩Ϊ˱ѧϰͻDavid Kwonһ˳(Neighbor Sequential)21ݣѧϰڼ䣬ÿ䵥ԪӦȨϵֻһΣEldracher˲Ӧ뼼22ݣCMACĺƽ
Chiang LinԭеCMACԭͽṹĻϣ˻ڹCMAC23ݣݳΪC-L CMAC˵CMACоһͻƣռɢ״̬ѡȡһĳ, ÿɢ״̬ÿɢ״̬мϸĹܣ洢ϢCMACɿ(Basis Function Network)CMAC뾶С24ݲ֮ͬ˻Ͳͬ⣬CMACÿھֲ(糬)Ըݵʱȷõĳ壬Ȼ㶨ЩϵĻϣΪCMACCMACͨؼϵȨϵʵֹ䣬ѧϰʱֻȨϵԽ֮ڵÿ΢ʱɼϵͳ΢ϢͳAlbus CMACɿǻΪ1磬ѧϰƽĵоٶȱAlbus CMAC󽵵ͣ
3.2CMACṹ
ڣ1) CMACڱƽӷԺʱͼ侫ʱҪ2) ͬһڸά⣬CMACάϸʱȨϵʮӴ󣬸ѧϰʵӦôѣˣоԱڶCMACṹ֣̽
Huang˲(Cascade)CMACṹ25ݣԸƵCMACӳľȣӦڴӡɫУ׼öάCMACĳֹһ൱һάCMACTham һH-CMAC(Hierarchical CMAC)ṹ26ݣЧؼٴ洢ռ䣬ڿѧϰӷԺLinLiCMACṹ27ݣÿСCMACϵͳռӼһάCMACΪɸάCMAC
4CMACѧϰ㷨
4.1Ȩϵ㷨
ParkMlizter28,WongSide29òͬķ֤Albus㷨ԣŷ˵оϴķڴͳCMACļ㷽ȫǡѧϰ㷨ֲģָĽѧϰ㷨Ϣ洢30ݣݵָAlbus㷨ѧϰǵȱݲ˸Ľ㷨31ݣLinKimӦѧϰ(Adaptive Critic Learning)3233, ѧϰṹCMACģ飺(Active)ģ(Critic)ģ飮ģõ֪ʶģѧϰƼChiangLinC-L CMACѧϰ㷨㷨23ݣ
4.2֯뾺ѧϰ
ڵCMACĿռǲģΪЧԪļ书ܣ֯3435ݵĸͨ֯ʵԪԼ㷨NieLinkensģCMAC(FCMAC)36ݣѧϰԪ(Association Cells)ƥȾԪĻʤGaoȣ37ݡChowMenozzi38, Ԫḻ̌׼ȷʤߣʹԪλƶӶѧϰȣͬʱήķ
5CMAC񾭿
񾭿ָڿϵͳвһ߶ԾȷĸӵķԶнģ䵱Ż㣬ϵȣԼͬʱĳЩܵʵϣϵͳͳΪĿϵͳֿƷʽΪƣʵܿƵҪʽ
1960WidrowHoffȰԶоһоܼѵȡһЩչKilmer McCullochȸݼ׵״ṹԭKMBģͲӦİ޵¼ƻ1964Widrow Smith˻Ӧõӣ60ĩ80ڣһڵͳоɹ٣80˹оĸպͷչ񾭿ƵоҲʮֻԾ
20оѽʾԵĽṹ͹ʵΪһʵϣϵͳϱֱۺ˵Ŀƣ˵ɵôӣԲҪ뻷ĶѧģͣҲκ΢ַ̣ϵͳֶԲȷӾȷͽĿǴͳƷԴﵽģڿӦȺȷչ񾭿ƷӦڹ̿ƣ3941У䶯Ҫ棺1)ϵͳԽԽӣ2)ʵֵĿԽԽߣڿƽ磬кܴ缼ͼķչΪ񾭿ṩ˼һһЩʺڿƵԺЩԺ1) NNϢĲдͿʵʱƺͶѧƣ2)籾ʵķΪԿƴµϣʹƲԿɲڶĽģͣ3)ͨѵѧϰܹЩѧģͻԴܴĿƹ̣4)кǿӦϢۺܹͬʱĲͬ͵Ŀ룬Ϣ֮ĻԺ⣬ʵϢۺϴرʹڸϵͳϵͳͶϵͳĿƣ
ͳĻģ͵ĿƷʽǸݱضѧģͼԿϵͳָƿԿƹɼѧģǻרҾ֪ʶܽģƹ򣬹вȷԸӶģϵͨϵͳ仯ģϵϳɻÿӶϵͳƣֿƷʽʽ֪ʶص㣬粻ʽ֪ʶкǿıƽԺӳڿŵ㣮
񾭿ƵĶ壬ԿԿеã42ݣͼ2ʾɷΪ:


ͼ2ڿеʾͼ
1) ޣģ͵ĸֿƽṹг䵱Ԥpredictor43ݣ£Ķ̬ģͿɿ뼰״̬ռ䵽ռľ̬ӳ䣬ɶⲿݶԤ2) ڷϵͳֱӳ䵱ã3) ڴͳϵͳŻã4) ܿƷŻ㷨ģƣ5)רҿƼŴ㷨ںΪṩǲģŻģͼϵȣ
ڿеӦÿɷΪࣺ1)еѧϰ߽У֪ܱƽ⻬ķԺѧϰĶ̬ԣȻӦõضַڶԲ2)񾭿ʷãͨѧϰ߿ܣ44ݣ
CMACر񾭿ƣҪΪ1) ֲƽѧϰٶȿ죬ѧϰҪ2) о֣45ݣCMACģ߼رǻڹCMACһģ߼ϵͳռɿѡģӼḶ́ռϵĻɿģӼCMACѧϰ̾Ѱģ߼Ḷ́
5.1罨ģ4647
CMACΪܱƽϵͳģͣǻΪض涯ѧģ߽ͣıƽģͣ۹
״̬ռʽȫϵͳȫ̬ΪҲɸϵͳһ㵫ȴԷƵıʽ⣬ϵͳݺṩ˶ϵĺʽʱУԻع黬ƽģ(ARMA),ֲͨƷҲɸϵͳڷϵͳڷԻع黬ƽģ(NARMA)ȴҵһǡĲƷͳķԿϵͳʶۺоʵӦжڼѣ
֮£ⷽʾԵԽԣͨѧϰƽӳӦڷϵͳĽģʶԲܷģ͵ƣұڸʵֵѧϰ㷨Ϣȡ(Data mining)оԾӴȡ̵Ķ̬ϢӶϵͳĶ̬ģͣһʮҪϢȡ뼯ɣͨһνģɻÿĽԾӦΪԤϵͳṩϣ
ӦоCMACѧϰԻرʺ߽ģУ
5.24849
ڹ̿УһҪǣ๤ҵ̵ܹպͼƣԼ羫ĲƷŨȣ͹޵ľŨȵȣֻͨѧķõЩҪܴʱһΣЩֵܵõĹ̵ɢģ̵ĿƺͼܴѣΪ˽⣬ַһǼӲøɲƲԲıƶϿƣ߷ǣǰЧʵãͶʴҲнϴͺڿҪϸߵĳҪ һǹҵ̱ʷ֮ûбȻϵĳֳķԤϵͳҪƻԲƷĳɲʽԤǰɳƷȣԼٲƷĿʱ䣮ϵͳУԭϣǳɲģȻڲͬʱڣ໥ģҲ໥ģڽɲԤģʱܵõʷֻһɢݣ
Ļ˼ǸݱȽײĹ̸αƲɲĹҪģ;Եһ̣ģ͵ĽһKalman˲ۻϡ෽ģͲУѴصĹ̡羫ϵͳͺѲģͣԲͳƻع鷨ģҲɲûϸķģģģʽʶҲһ෽һǳҪķģͽķ編Ϊвмѧϰݴص㣬ɸݶֱӽģҪ̫֪ʶҿԽУѧϰ㷨ģ͵У漰У㷨ЧԺͿ⣮һЧȡУݵѡȡȡ㷨CMACĿѧϰصرʺ
ģʵһֻľ̬ģͣ CMAC綯̬ģ̬ģ:CMACΪ̬ģʱCMACǾж̬ԵУCMACΪӦһеĹCMACΪ̬ģʱǹĳһ룬CMACΪһµ̬Ҫ4棺1)βѡ2)ݴ3)ģ͵Ľ4)ģ͵УαѡǳҪΪɲҪҪЩαƶϳαѡ3Ҫ棺1)͵ѡѡķӼָ磬Ʒһ̤¶ȣзӦеĲƷ÷Ӧܱ¶ȣ2)ѡУŶαĿǺȷ⣬ʵӦʱҪϵͳĻҪȷСĿȻϾӶαĿ3)ѡȡɲȡֵֽķҲɲùҵƷȷ
5.3񾭿ϵͳ
ƺʵĿ漰3棺1 )ѡǡĿṹеĲҪȷеҪͨѧϰõ2)ñضйϢഫͳĿʱ֪֪΢ַ̣༫˵ǶԱض˽٣ɵõⲿݣ3)ĳֿϵͳָ꣮
CMAC񾭿ĲȷһĽṹάռȣѧϰ㷨ĲôʵȨֵṹͿѧϰңָض
Ŀǰ񾭿Ƶķͳһı׼ݷΪ1) רϵͳƣרϵͳڱ֪ʶ߼ϷӸԵƣͻøõĿЧ 2) 涯̬ƣϵͳ涯̬ģʹڣͨѵ磬ϵͳ涯̬ģͣ3) CMACģ߼ƣ5051ݣCMACģ߼ģCMAC(FCMAC), ֯;ѧϰܣģ߼ʹCMACģ߼ƽнϺõĿЧ4)ģͲοӦƣ5253ݣͬģͲοӦϣ͹ģͲοӦƣϵͳṹʽϵͳģͲοӦͬģֻͨCMACıʶģͣ5)Уƣ54ݣķԺӳʹУϵͳг䵱δ֪ϵͳƽ6)ģƣ55ݣ磬һ񾭿һ״̬񾭿ǰһ˲״̬ڳֱƽĶ̬ģͣ񾭿ֱѧϰض涯̬ģͣԳ״̬ģͣڲģͣΪѧϰ󣬼ӵѧϰض涯̬ԣ7)Ԥƣ5657ݣͨѧϰʵϵͳݣCMACɱƽضĶ̬ԣڿʱCMACΪԤԤδӶʵԤƣַʽһۺCMACԤֱƿֻһCMACCMACCMACԤṹ˫CMACṹַϴ󣮴⣬һЩĿƷĳϵͳĿ⣬ϵͳķϵͳ58ݣBrunowskiiϵͳ59ݵȣ
6CMAC񾭿6061
оԱݴͳCMACĽṹ㷨򹤳̵ӦʵCMACĽṹѧϰ㷨ģȣЩĿӦɻƶCMAC۵оӦõƹ㣮
7Ҫ
һֺõ۱ѭ׼1) Perform network design task;2)ѧϰ³ԣRobustness in learning3)ѧϰԣQuickness in learning 4)ѧϰЧԣEfficiency in learning5)ѧϰԣGeneralization in learning
ڶ̶̵ļ񾭿Ƶо۴۵ӦöȡϲĽչҲ뿴Ƕϵͳо˽⻹Ƿȱʹõʽģ۴ӽṹģʵļģ⣬񾭿ƵоԭʼϵͳϵսԵδӦóɹʵҲһƹ㣮
CMAC񾭿оĿڽͼ£Ѱҿȫѧϰ㷨ԼCMACѰҾкǿ³ԵȫȶĿƲԣʵӦʵ֣ĿǰCMAC񾭿Ƶо״¼ؽ
1) ά䵥Ԫ(Ȩϵ)ľ⣮ǰļͼķȱݷеķȱݣ2) ϵͳ³ԣĿǰ񾭿ƵооûиŻԤʱѧϰ㷨ϵͳȶԣ³ԣʵϣ񾭿Ƶ³ǼҪģرڹ̿Уͬʱ³Ҳ񾭿Ƶ㣮3) Чѧϰ㷨о4) CMACάĵȷCMACƽϵͳʱCMACʵϵͳʷʷǰɣͼ3ʾȻԲͬʵϵͳʷݵĿͬCMACάͲͬǡȷCMACάTDLĳȣֵоĿ⣮ 5) 񾭿ϵͳܿԡܹԼȶԣ6) 񾭿㷨оرоʺֲʽмصĿѧϰ㷨Գģѧϰ㷨Ӧ񾭿רоƬ


ͼ3CMACļѧϰϵͳ
޷϶񾭿ۼӦоʲôͻԳɹԿ϶ǣ񾭿Ǻϣо
߼飺31꣬ʿоΪ񾭿ơģơӦƵоӦã
ۻݺף63꣬ڣоΪģͻŻơܿоӦã
ߵλϺͨѧԶϵ200030
ο
1 Albus,J S. Data Storage in the Cerebellar Model Articulation Controller (CMAC).Transaction of ASME, J. Dynam. Syst. Meas. Control, 1975, 97: 228233
2 Eccles J C. The Cerebellum as Computer: Patterns in Space and Time, J. Physical, 1973, 229:132
3 Commuri S, Jagannathan S, Lewis F L. CMAC Neural Network Control of Robot Manipulators. Journal of Robotic Systems. 1997, 14: 46582
4 Mller W T. Real-time Application of Neural Networks for Senser-based Control of Robots with Vision,IEEE Tans. Sys. ,Man, Sybern., 1989,19: 825831
5¡ݣ̣࣬߾.ѧ,1996,47(3):376-80
6 Pinsopon Unnat, et al. Real-time CMAC Neural Network Control of a Piezo-electric Actuated Toolpost with Disturbance Decoulped Observer. Proceedings of the 1996 ASME International Mechanical Engineering Congress and Exposition. ASME, New York, USA. 1996:325331
7 Koo Keun-mo, kim Jong-Hwan. CMAC Based Control of Nonlinear Mechical System. Proceedings of the 1996 IEEE 22nd International Conference on Industrial Electronics, Control, and Instrumentation, IECON. Taipei,Taiwan. 19541959
8 An P Edgar, Harris Chris J. Intelligent Driver Warning System for Vehicle Collision Avoidance. IEEE Tran. Syst. Man and Cybern.Part A Systems and Humsans ,1996,26(2):254261
9 Kawamoto Isao, Iiguni Youji, Adachi Norihiko. Design of Cerebellar Aodel Arithmetic Computer with Adaptive Resolution and its Application to Nonlinear Signal Processing. Electronics & Communication in Japan, Part3:Fundamental Electronic Science. 1995,78(9): 3139
10Shiraishi Hitosh, Ipri Susan L, Cho Dong-il D. CMAC Neural Network Controller for Fuel-injection Systems. IEEE Trans on Control Systems Technology, 1995, 3(1): 3238
11Jar-Shone, et al. A Fuzzy CMAC Model for Color Reproduction. Fuzzy Sets and Systems, 1997, 91:5368
12Chen X H, Gao F, Qian J X. Adaptive Control Based on RBF Networks. Proceedings of the 35th Conference on Decision and Control, Kobel, Japan, 1996: 38103815
13ˣ˹.廪ѧ磬1996
14Chien C J. A Discrete Iterative Learning Control of Nonlinear Time-varying Systems. Proceedings of the 35th Conference on Decision and Control, Kobel, Japan, 1996: 305661
15Ker Jar-Shone et al. Hardware Realization of Higher-order CMAC Model for Color Calibration. Proceedings of the 1995 IEEE International Conference on Neural Networks. 1995:165661
16Ker Jar-Shone et al. Hardware Realization of Higher-order CMAC Model for Color Calibration. Proceedings of the 1995 IEEE International Conference on Neural Networks. 1995:165661
17Penterson J K. Software Design for Neurocontrol Code. Journal of Intelligent Material Systems and Structures, 1994,5:17285
18An P E et al. Associative Memory Neural Networks: Adaptive Modelling Theory, Software Implementations and Graphical user Interface. Engineering Applic. Artif. Intell. 1994, 7(1):121
19ޡңлıع⣮CMACѧϰԵоԶѧ1997,23(4): 455461
20Wang Zi-Qin, Schiano Jeffrey L, Ginsberg Mark. Hash-coding in CMAC Neural Networks. Proceedings of the 1996 IEEE International Conference on Neural Networks, ICNN. Washington, DC, USA, 1996:16981703
21David E, Kwon S. Neighbor Sequential and Random Training Techniques for CMAC. IEEE Tran. Neual Networks,1995,6(1):196202
22Eldracher M, Staller A, Pompl R. Adaptive Encoding Strongly Improves Function Approximation with CMAC. Neural Computation, 1997, 9(2): 40317
23Chiang&Lin. CMAC with General Basis Functions, Neural Networks. 1996, 9(7):11991211
24Qinghua Zhang. Using Wavelet Network in Nonparametric Estimation. IEEE Trans. Neural Networks. 1997, 8(2): 227236
25King-Lung Huang, Shu-cheng Hstesh, Hsin-Chia Fu. Cascade-CMAC Neural Network Application on the Color Scanner to Printer Calibration. Proceedings of the 1997 IEEE International Conference on Neural Networks. 1997:1015
26Tham C K. Hierarchical CMAC Architecture for Context Dependent Function Approximation. Proceedings of IEEE Conference on Neural Networks, Washington, DC, USA, 1996:629634
27Lin C S, Li C K. A New Neuronetwork Structure Composed of Small CMACs. Proceedings of IEEE Conference on Neural Networks, Washington, DC, USA, 1996:17771783
28Parks P C, Militzer J. Convergence Properties of Associative Memory Storage for Learning Control System. Automation and Remote Control. 1989:25486
29Wong Y, Sideris A. Learning Convergence in the Cerebellar Model Articulation Controller. IEEE Trans. Neural Networks,1992,3(6):115121
30Wong Y, Sideris A. Learning Convergence in the Cerebellar Model Articulation Controller. IEEE Trans. Neural Networks,1992,3(6):115121
31ݣӿСģĽ㷨о,Զѧ,1997,23(4):482487
32C S Lin, H Kim. CMAC-based Adaptive Critic Self-learning Control. IEEE Trans. Neural Networks,1991,2(5):53033
33C S Lin, H Kim. Selection of Learning Parameters for CMAC-based Adaptive Critic Self-Learning Control. IEEE Trans. Neural Networks,1995,6(3):64247
34Kohonen T. The Self-organizing Map. Proceedings of the IEEE Special Issue on Neural Networks I Theory and Modelling,1990:146480
35Wann C D et al. A Comparative Study of Self-organizing Clustering Algorithms Dignet and ART2. Neural Networks, 1997,10(4): 73754
36Nie J, Linkens D A. FCMAC: a Fuzzied Cerebellar Model Articulation Controller with Self-organizing Capacity. Automatica, 1994,30(4):655644
37Gao Xiaozhi, Wang Changhong, Gao X M, Ovaska Seppo J. New CMAC Neural Network Model with Adaptive Quantization Input Layer. Proceedings of ICSP'96, Beijing, China, 1996:14171420
38Chow M, Ymenozzi A. A self-organized CMAC Controller. Proceedings of IEEE Conference on Industrial Technology, Guangzhou, China, 1994:6872
39¡,,̣߾.ѧ,1996,47(3):37680
40He H et al. The Self-learning CMAC Control of the Oil Temperature on Hydraulic System. Proceedings of IAS'95, Shenyang, China, 15991604
41Quek C, Ng P W. Realisation of Neural Network Controllers in Integrated Process Supervision. Artificial Intelligence in Engineering. 1996, 10(2):135142
42Mulkul Agarwal. A Systematic Classification of Neural-net-work-based Control. IEEE control systems, 1997:7593
43Yonghong Tan, Achiel van Cauwenberghe. Nonlinear One-step-ahead Control using Neural Networks:Control Straegy and Stability Design. Automatica, 1996, 32(12):17011706
44Xu J X. Direct Learning of Control Efforts for Trajectories with Different Times Scales. Proceedings of the 35th Conference on Decision and Control, Kobel, Japan, 1996:250102
45ۻݺףһֻģCMACѧϰߣ¼
46Chien Chere Cheah, DanWei Wang. A Model Reference Learning Control Scheme for a Class of Nonlinear Systems. Int. J. Control.1997,66(2):271287
47G P Liu et al. Stable Seqential Identification of Continuous Nonlinear Dynamical Systems by Growing Radial Basis Function Networks. Int. J. Control.1996,65(1):5369
48Wang Xudong, Shao huihe, Lou Rongfu. Designing a Soft Sensor for a Distillation Collumn with the Fuzzy Radial Basis Function Neural Network, 35th IEEE Conference on Decision and Control, Kobe, 1996:17141719
49񶫣RBF缰ڿеӦãʿѧλģ1997
50Shi Weixiang, Zhou Fuzhang. CMAC Neural Network Fuzzy Control. Proceedings of the 1996 IEEE IECON. 22nd International Conference on Industrial Electronics, Control, and Instrumentation. 1996:5849
51ߣ־CMACģ缰ڿеӦã廪ѧѧ199636(5):1723
52Chen Fu-Chuang, Hassan K Khilil. Adaptive Control of a Class of Nonlinear Discrete Time Systemss using Neural Networks. IEEE Trans. Automatic Control, 1995, 40(5):791801
53Hu Y, Fellman, Ronald D. Implementation Effient Learning Algorithm for Adaptive Control using Associative Content Addressable Memory. IEEE Tran. Syst. Man and Cybern, 1995,25(4):70409
54Ahmed M S, M Farooq. Neural-net-based Direct Self-tuning Control of Nonlinear Plants. INT. J. CONTROL, 1997,66(1): 85-104.Computers chem. Engng., 1992, 16(12):10391057
55Nahas E P et al. Nonlinear Internal Model Control Strategy for Neural Network Models. Computers Chem. Engng., 1992, 16(12):10391057
56Lin, Chun Shin Kim, Hyongsuk Selection of Learning Parameters for CMAC-based Adaptive Critic Learning. IEEE Transactions on Neural Networks. 1995, 6:6427
57Tan Y, Cauwenberghe A V. Nonlinear One-step-ahead Control using Neural Networks: Control Strategy and Stability Design. Automatica, 1996, 32(12):17011706
58Lee S W, Kim J H. Control of Systems with Deadzones using Neural Network Based Learning Controller. Proceedings of IEEE Conference on Neural Networks, 1994,253538
59Commuri S, Lewis F L. CMAC Networks for Control of Nonlinear Dynamical System: Structure, Stability and Passivity. Automatica,1997,33(4):63541
60P E An et al. Associative Memory Neural Networks: Adaptive Modelling Theory, Software Implementations and Graphical user Interface. Engng Applic. Artif. Intell. 1994,7(1):121
61James K. & Peterson. Software Design for Neuralcontrol Code. Journal of Intelligent Material Systems and Structures. 1994, 5: 172185
62Farrell J A. Persistence of Excitation Conditions in Passive Learning Control. Automatica, 1997,33(4):699703
63Chen Fu-chuang, Chang Chih Horng. Practical Stability Issues in CMAC Neural Network Control Systems. IEEE Trans. On Control Systems Technology. 1996,4(1): 8691
1998-04-22ո
