Neural tracking trajectory of discrete-time nonlinear systems based on an exponential sliding mode algorithm

Abstract

For a class of discrete-time nonlinear systems, this paper proposes a discrete-time sliding mode control algorithm with an exponential term that acts as a variable gain to improve the convergence to a bounded region around the origin for the state variable. To compensate for unknown, but bounded disturbances, it has been employed a discontinuous function through a discrete-time integral action. A discrete-time neural network has also been employed to identify such nonlinear system with a block controllable structure. Finally, the proposed controller and the neural network algorithms are applied to a discretised direct current (DC) motor to follow a desired admissible trajectory. Numerical simulations show the effectiveness of the proposed control law.