An introduction to the use of neural networks in control systems

Abstract

SUMMARYThe purpose of this paper is to provide a quick overview of neural networks and to explain how they can be used in control systems. We introduce the multilayer perceptron neural network and describe how it can be used for function approximation. The backpropagation algorithm (including its variations) is the principal procedure for training multilayer perceptrons; it is briefly described here. Care must be taken, when training perceptron networks, to ensure that they do not overfit the training data and then fail to generalize well in new situations. Several techniques for improving generalization are discussed. The paper also presents three control architectures: model reference adaptive control, model predictive control, and feedback linearization control. These controllers demonstrate the variety of ways in which multilayer perceptron neural networks can be used as basic building blocks. We demonstrate the practical implementation of these controllers on three applications: a continuous stirred tank reactor, a robot arm, and a magnetic levitation system.

Design of conventional and neural network based controllers for a single-shaft gas turbine

“…Rowen and Housen (1983) investigated GT airflow control for optimum heat recovery and its advantages at gas turbine part-load conditions. Hagan et al ( (2002Hagan et al ( ( ), (1999) presented an overview of neural networks and their applications to control systems.…”

Analysis of volumetric swelling and shrinkage of heat treated woods: Experimental and artificial neural network modeling approach

“…The second phase is known as the validation phase. This phase is carried out to generalize the established network (Hagan et al 2002). The last phase is the testing phase that resembles the validation phase except in one difference.…”

“…The last phase is the testing phase that resembles the validation phase except in one difference. The validation phase is considered as a criterion to end the training phase, while testing phase is performed to measure the performance of the trained and validated network (Hagan et al 2002, Avunduk et al 2014). Thus, it is possible to properly evaluate the performance of the established model (May et al 2010).…”

Neural and Genetic Control Approaches in Process Engineering

“…The final network structure selected for the neural composition estimator was a 4-25-2 net, trained using the Levenberg-Marquardt algorithm (Hagan et al, 2002), with a hidden layer configuration selected after a trial and error process and input layer determined by the PCA based algorithm for selection of the secondary variables previously exposed.…”