Real-time application of neural model predictive control for an induction servo drive

This paper deals with a real-time implementation of a neural net controller and its application to an induction machine. The controller used within this framework is the model-based predictive control structure (MPC) with a neural network model of the process dynamics. The hardware platform, a multi-transputer system, is described as are the implemented algorithms. Some results show the benefits of the neural network controller compared to the field oriented control structure which is the most common structure used to control induction machines