Neural network integration in control system structures: a case study towards intelligent control by neural network integration

Neural networks techniques have been found to be useful for controlling highly uncertain, nonlinear and complex systems. The possibility of applying neural networks for modelling of unknown functions in dynamic environments has been demonstrated by several empirical studies. Allowing unknown nonlinearities in the system to be controlled we extend the class of uncertain systems that adaptive control techniques can be applied to. In the paper an integration of neural networks in the control systems is considered. Neural network application schemes are described. Some advantages of neural modelling and control are presented.