Process control by on-line trained neural controllers

Abstract

The question of how to perform online training of multilayer neural controllers in order to reduce the training time is addressed. First, based on multilayer neural networks, structures for a plant emulator and a controller are described. Basic control configurations are briefly presented, and new online training methods, based on performing multiple updating operations during each sampling period, are proposed and described in algorithmic form. One method, the direct inverse control error approach, is effective for small adjustments of the neural controller when it is already reasonably trained; another, the predicted output error approach, directly minimizes the control error and greatly improves convergence of the controller. Simulation and experimental results using a simple plant show the effectiveness of the proposed control structures and training methods.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>