A General Learning Scheme for CMAC-based Controller

Cerebellar model articulation controller (CMAC) is a powerful tool for nonlinear control applications. However, it yet lacks an adequate learning scheme. It is found that, with the existing learning scheme, if a complicated learning algorithm is not used, CMAC can destabilize a system that is otherwise stable. Oscillations resulting from the interaction between CMAC and the classical controller were found to contribute to the instability. This paper presents a new CMAC learning scheme that models plant's characteristics based on closed loop errors instead of the original input-output pairs. In this scheme, memory space of the CMAC is partitioned into two parts. One is for dynamic control, in which dynamic information is stored. Another is for steady state control, in which steady state information is adaptively updated for smooth control. Relationship between the two parts of the space is discussed and specified for a stable control. Simulation results on a typical nonlinear plant model and a real electrohydraulic servo system using the proposed scheme demonstrate that the oscillations are eliminated and stable control is obtained. The new scheme demonstrates superior tracking performance, noise rejection property and good robustness.

[1]  James S. Albus,et al.  I A New Approach to Manipulator Control: The I Cerebellar Model Articulation Controller , 1975 .

[2]  R. Guenther,et al.  VS-ACC applied to a hydraulic actuator , 2000, Proceedings of the 2000 American Control Conference. ACC (IEEE Cat. No.00CH36334).

[3]  James S. Albus,et al.  New Approach to Manipulator Control: The Cerebellar Model Articulation Controller (CMAC)1 , 1975 .

[4]  Samuel F. Asokanthan,et al.  CMAC based controller for hydro-mechanical systems , 2001, Proceedings of the 2001 American Control Conference. (Cat. No.01CH37148).

[5]  Sabri Cetinkunt,et al.  Adaptive learning algorithm for Cerebellar model articulation controller , 2000 .

[6]  Li Yunhua,et al.  Development of Hybrid Control of Electrohydraulic Torque Load Simulator , 2002 .

[7]  Mario Sznaier.  Robust control of dynamic systems using neuromorphic controllers: a CMAC approach , 1992, [1992] Proceedings of the 31st IEEE Conference on Decision and Control.

[8]  Cling-An Lin,et al.  On a new CMAC control scheme, and its comparisons with the PID controllers , 2001, Proceedings of the 2001 American Control Conference. (Cat. No.01CH37148).

[9]  Fu-Chuang Chen,et al.  Practical stability issues in CMAC neural network control systems , 1994, Proceedings of 1994 American Control Conference - ACC '94.

[10]  James S. Albus,et al.  Data Storage in the Cerebellar Model Articulation Controller (CMAC) , 1975 .

[11]  Frank L. Lewis,et al.  Feedback Linearization using CMAC Neural Networks , 1998, Autom..

[12]  Sabri Cetinkunt,et al.  Hydraulic actuator control with open-centre electrohydraulic valve using a cerebellar model articulation controller neural network algorithm , 1999 .

[13]  Frank L. Lewis,et al.  Feedback linearization using CMAC neural networks , 1996, Proceedings of 35th IEEE Conference on Decision and Control.

[14]  Hung-Ching Lu,et al.  Robust CMAC control schemes for dynamic trajectory following , 2002 .

[15]  Chao He,et al.  Learning Convergence of CMAC Algorithm , 2004, Neural Processing Letters.

[16]  W. Thomas Miller,et al.  Cerebellar model arithmetic computer , 1996 .

[17]  Frank L. Lewis,et al.  CMAC neural networks for control of nonlinear dynamical systems: Structure, stability and passivity , 1997, Autom..

[18]  Kai Keng Ang,et al.  Improved MCMAC with momentum, neighborhood, and averaged trapezoidal output , 2000, IEEE Trans. Syst. Man Cybern. Part B.

[19]  Chien-Chung Lin,et al.  Improved CMAC neural network control scheme , 1999 .

[20]  Filson H. Glanz,et al.  Application of a General Learning Algorithm to the Control of Robotic Manipulators , 1987 .