ADAPTIVE-CONTROL OF DISCRETE-SYSTEMS USING NEURAL NETWORKS

A neural network controller which is used for controlling unknown disc rete-time DARMA systems is described. A two-layered neural network is used to estimate the unknown plant dynamics. The well known Widrow-Hof f delta rule is used as the learning algorithm for this network, to mi nimise the difference between the plant actual response and that predi cted by the neural network. The control law is generated online using a second two-layered neural network, so that the plant output is broug ht to a desired reference signal. It is proved that the control object ive is achieved by the closed-loop system and that the system remains closed-loop stable. Some simulation examples are also presented to eva luate the design.