An application of SMC theory for experimental learning control of robotic manipulators

Complexity of learning dynamics constitutes a prime difficulty in online neurocontrol schemes involving gradient computations in parameter update rules. This is because such complexities can make closed loop system sensitive to uncertainties. In this paper, we discuss a learning control approach, which is based on the sliding mode control (SMC) techniques instead of gradient computations. Due to properties of SMC, learning process becomes robust to uncertainties. In order to test the control scheme, we have chosen a robotic manipulator as the test bed. Experimental results show that the control approach achieves a good tracking performance.

[1]  T.,et al.  Training Feedforward Networks with the Marquardt Algorithm , 2004 .

[2]  Stefen Hui,et al.  Application of feedforward neural networks to dynamical system identification and control , 1993, IEEE Trans. Control. Syst. Technol..

[3]  D. Marquardt.  An Algorithm for Least-Squares Estimation of Nonlinear Parameters , 1963 .

[4]  Mohammad Bagher Menhaj,et al.  Training feedforward networks with the Marquardt algorithm , 1994, IEEE Trans. Neural Networks.

[5]  Stanislav V. Emelyanov,et al.  Variable-Structure Control Systems , 1995 .

[6]  Xinghuo Yu,et al.  Sliding Mode Control of a Three Degrees of Freedom Anthropoid Robot by Driving the Controller Parameters to an Equivalent Regime , 2000 .

[7]  Kenneth Levenberg.  A METHOD FOR THE SOLUTION OF CERTAIN NON – LINEAR PROBLEMS IN LEAST SQUARES , 1944 .

[8]  U. Yildiran,et al.  Neural network based control of a cement mill by means of a VSS based training algorithm , 2002, Industrial Electronics, 2002. ISIE 2002. Proceedings of the 2002 IEEE International Symposium on.

[9]  Antônio de Pádua Braga,et al.  Sliding mode algorithm for training multilayer artificial neural networks , 1998 .

[10]  Eliezer Colina-Morles,et al.  A sliding mode strategy for adaptive learning in Adalines , 1995 .

[11]  Frank Fallside,et al.  An adaptive training algorithm for back propagation networks , 1987 .

[12]  Stanislaw H. Zak,et al.  The adaptation of perceptrons with applications to inverse dynamics identification of unknown dynamic systems , 1991, IEEE Trans. Syst. Man Cybern..