Artificial Neural Networks
Published Online: 15 SEP 2009
Copyright © 2009 John Wiley & Sons, Ltd. All rights reserved.
Encyclopedia of Structural Health Monitoring
How to Cite
Reed, S. 2009. Artificial Neural Networks. Encyclopedia of Structural Health Monitoring. .
- Published Online: 15 SEP 2009
Artificial neural networks (ANNs) are able to learn from experience, generalize from examples, and identify underlying information from within noisy data. These characteristics, and the explosion in ANN techniques over the past 15 or 20 years, have led to an ever-increasing role for ANNs within structural health monitoring (SHM) systems. Within this article, the development of ANNs and the basic principles behind their functionality, training, and deployment are described. Emphasis is placed upon the multilayer perceptron (MLP), as the workhorse of ANN methods; however, other techniques, including radial basis functions (RBF) and self-organizing feature maps (SOFMs), are also detailed. Examples of the use of ANNs within SHM systems are presented and reference has been made to a wide range of texts throughout this article.
- artificial neural networks;
- novelty detection;
- damage detection;
- structural usage monitoring