Real-time adaptive neural control of a class of nonlinear systems

An adaptive neural control scheme is tested here by means of real-time control experiments developed for controlling a class of linear and nonlinear systems. The adopted neural control scheme has many stability and convergence properties, which were analyzed from a theoretical point of view in a previous work of the authors. The experimental results presented here extend the scope of the neural control design as it was originally formulated, and demonstrate that it can be used for controlling both linear and nonlinear plants. The experiments also prove that the proposed neural control design performs well, not only in theory, but in real tests as well. (C) 1999 Elsevier Science Ltd. All rights reserved.