Computer-Aided Civil and Infrastructure Engineering 21 (2006) 232–241 Structural Health Monitoring via Measured Ritz Vectors Utilizing Artificial Neural Networks Heung-Fai Lam ∗ Department of Building & Construction, City University of Hong Kong, Hong Kong SAR Ka-Veng Yuen Faculty of Science and Technology, University of Macau, P.O. Box 3001, Macau SAR & James L. Beck Division of Engineering and Applied Science, California Institute of Technology, Pasadena, CA 91125 Abstract: A pattern recognition approach for structural health monitoring (SHM) is presented that uses damage- induced changes in Ritz vectors as the features to charac- terize the damage patterns defined by the corresponding locations and severity of damage. Unlike most other pat- tern recognition methods, an artificial neural network (ANN) technique is employed as a tool for systematically identifying the damage pattern corresponding to an ob- served feature. An important aspect of using an ANN is its design but this is usually skipped in the literature on ANN-based SHM. The design of an ANN has significant effects on both the training and performance of the ANN. As the multi-layer perceptron ANN model is adopted in this work, ANN design refers to the selection of the num- ber of hidden layers and the number of neurons in each hidden layer. A design method based on a Bayesian prob- abilistic approach for model selection is proposed. The combination of the pattern recognition method and the Bayesian ANN design method forms a practical SHM methodology. A truss model is employed to demonstrate the proposed methodology. ∗ To whom correspondence should be addressed. E-mail: paullam@ cityu.edu.hk. 1 INTRODUCTION The safety of structures, such as buildings and bridges, is of serious public concern. Reliable and efficient struc- tural health monitoring (SHM) methods can play an important role in addressing this issue. Changes in dy- namic characteristics due to localized loss of stiffness are frequently employed as a means of structural dam- age detection. Due to advances in sensor technologies, instruments have been developed, such as laser Doppler vibrometers and shearographs, which provide accurate dynamic structural response measurements. However, the development of methodologies and algorithms for extracting useful information from the measured data for the purpose of SHM has not yet reached a stage of successful real applications. The objective of this work is to develop a practical SHM methodology that is based on a pattern recognition approach and artificial neural networks (ANNs). The proposed SHM methodology makes use of the measured load-dependent Ritz vectors, which have been shown in some situations to be more sensitive to stiffness loss accompanying structural damage than natu- ral frequencies and mode shapes (Cao and Zimmerman, 1997a,b; Sohn and Law, 2001). The idea of using Ritz C  2006 Computer-Aided Civil and Infrastructure Engineering. Published by Blackwell Publishing, 350 Main Street, Malden, MA 02148, USA, and 9600 Garsington Road, Oxford OX4 2DQ, UK.