Neural networks for control of industrial processes

In this contribution, neural concepts and methods for control an their use in industrial applications are discussed and illustrated. Even though there exist numerous conventional approaches for solving control tasks, their realization in practice frequently proves to be very difficult. The author pursues several approaches to using neural networks in the context of nonlinear control tasks. In identification, networks are trained to model the dynamics of an unknown nonlinear plant. The model provides the basis for controller design or system diagnosis. In robot control, networks are trained to model the inverse dynamics of the robot. The inverse model is used to linearize the system which is then accessible for the well-established tools of linear control theory. In another context, a network is used as a nonlinear trainable controller