Content area
Abstract
A neural network (NN) tracking control for the known affine in the inputs continuous time nonlinear dynamic systems is investigated. For this purpose, steady states, stability and convergence to steady states of state variable trajectories are analyzed. It is depicted that the steady states of state variable trajectories of the system exist, and they are unique. On a global scale, the state variable trajectories are stable and convergent to the steady states. The tracking control provides finite convergence time of the system state variable trajectories to steady states and decreasing tracking errors. The convergence time is constrained by altering the learning rate parameter. Sliding modes of the state variable trajectories are analyzed. Computer simulations of the controller operation confirming theoretical derivations and illustrating its high performance are provided. The NN can be used for efficient tracking control in real time the known affine in the inputs continuous-time nonlinear systems with known internal dynamics.
