Motion control with deadzone estimation and compensation using GRNN for TWUSM drive system

The traveling wave ultrasonic motor (TWUSM) drive has a strong nonlinearity, which varies with driving conditions and possesses variable deadzone in the control input associated with the phase difference of applied two-phase AC driving voltages. This deadzone is a problem as an accurate positioning actuator for industrial application, and it is important to eliminate the deadzone in order to improve the control performance. To overcome the above problems, a new motion control scheme with deadzone estimation and compensation using generalized regression neural network (GRNN) is proposed in this paper to improve the control performance of the TWUSM drive system. One of them is to approach the nonlinearity deadzone part of the TWUSM drive system, and another one with extension form is to approach the inverse of nonlinearity deadzone to realize the dynamic compensation. Once the nonlinearity of deadzone is compensated exactly by its inverse compensation, the whole TWUSM drive system can be treated as linear, and the estimated parameters of the linear system can be adopted to design a controller for a desired reference command.

[1]  Toshiiku Sashida,et al.  An Introduction to Ultrasonic Motors , 1994 .

[2]  Marios M. Polycarpou,et al.  Learning and convergence analysis of neural-type structured networks , 1992, IEEE Trans. Neural Networks.

[3]  Chun-Jung Chen,et al.  A Novel Driver with Adjustable Frequency and Phase for Traveling-Wave Type Ultrasonic Motor , 2007, Second International Conference on Innovative Computing, Informatio and Control (ICICIC 2007).

[4]  K. T. Chau,et al.  Speed Control of Traveling-wave Ultrasonic Motors Using a Practical Modeling Approach , 2007 .

[5]  Chun-Yi Su,et al.  Robust adaptive control of a class of nonlinear systems with unknown dead-zone , 2004, Autom..

[6]  W. Zhonghua,et al.  Robust adaptive deadzone compensation of DC servo system , 2006 .

[7]  Hilmi Berk Celikoglu,et al.  Application of radial basis function and generalized regression neural networks in non-linear utility function specification for travel mode choice modelling , 2006, Math. Comput. Model..

[8]  Mi-Ching Tsai,et al.  Neuro-fuzzy speed control of traveling-wave type ultrasonic motor drive using frequency and phase modulation. , 2008, ISA transactions.

[9]  Matthew C. Turner.  Actuator deadzone compensation: theoretical verification of an intuitive control strategy , 2006 .

[10]  Donald F. Specht,et al.  A general regression neural network , 1991, IEEE Trans. Neural Networks.

[11]  Rastko R. Selmic,et al.  Neural network control of a class of nonlinear systems with actuator saturation , 2004 .

[12]  Gang Tao,et al.  An adaptive dead-zone inverse controller for systems with sandwiched dead-zones , 2003 .

[13]  Tomonobu Senjyu,et al.  Position control of ultrasonic motors using MRAC and dead-zone compensation with fuzzy inference , 2002 .

[14]  Prashanth Krishnamurthy,et al.  Robust Adaptive Control of a Class of Nonlinear Systems , 2003 .

[15]  Zhao Chunsheng,et al.  Precise position control of ultrasonic motor using fuzzy control , 2004, IEEE Ultrasonics Symposium, 2004.

[16]  Chi Yung Yen,et al.  Thin-disc piezoceramic ultrasonic motor. Part II: system construction and control. , 2003, Ultrasonics.