E. Irigoyen, J.B. Galván, and M.J. Pérez-Ilzarbe (Spain)
Neural Networks, Optimal Control, Constraints, Nonlinear Systems
In this work, two optimal neural controller structures are studied in a noisy control scheme with nonlinear systems. For training these structures a modified Backpropagation algorithm is proposed. Moreover, a set of constraints in the training algorithm and in the control scheme are introduced. Two nonlinear systems with complicated dynamics have been proved in tests that show the good behaviour of the controllers.
Important Links:
Go Back
