Experimental Assessment of Vibration–Based Time Series Methods for Structural Health Monitoring Fotis P. Kopsaftopoulos and Spilios D. Fassois Stochastic Mechanical Systems & Automation (SMSA) Laboratory Department of Mechanical & Aeronautical Engineering University of Patras, GR 265 00 Patras, Greece E-mail: {fkopsaf,fassois}@mech.upatras.gr Internet: http://www.smsa.upatras.gr ABSTRACT This work aims at the experimental assessment of vibration–based time series methods for Structural Health Monitoring (SHM). The methods are classified as non– parametric and parametric, while their features and operation are discussed. Further- more, their performance characteristics and ease of use are assessed and compared. The application and comparative assessment of the methods through experiments on a laboratory aluminum truss structure is presented. The results of the study confirm the high potential and effectiveness of the statistical time series methods for SHM. INTRODUCTION Vibration–based statistical time series methods for damage detection and identi- fication utilize random excitation and/or vibration response signals, along with sta- tistical model building and decision making tools, for inferring the health state of a structure [1–4]. They offer a number of important advantages, including no require- ment for physics–based or finite element type models, no requirement for complete modal models, effective treatment of uncertainties, and statistical decision making with specified performance characteristics. In spite of the above, the literature on vibration–based time series methods for SHM remains relatively sparse, and, in par- ticular, no application studies that assess and compare the various methods are avail- able. 4th European Workshop on Structural Health Monitoring (SHM 2008) Cracow, Poland, July 2008 The goal of this study is to contribute to filling this gap by presenting the applica- tion and comparative assessment of a number of statistical time methods to a labora- tory aluminum truss structure. The damages considered correspond to the loosening of bolts connecting certain of the truss elements. Random force excitation is pro- vided via an electromechanical shaker, while the vibration responses are measured at various positions via lightweight strain gauges. Two non–parametric methods (Power