353 SDSS’Rio 2010 STABILITY AND DUCTILITY OF STEEL STRUCTURES E. Batista, P. Vellasco, L. de Lima (Eds.) Rio de Janeiro, Brazil, September 8 - 10, 2010 STRUCTURAL DAMAGE ASSESSMENT USING THE DIFFERENTIAL EVOLUTION AND THE ANT COLONY OPTIMIZATION TECHNIQUES Genasil F. dos Santos 1 , José Guilherme S. da Silva 2 , Francisco J. da C. P. Soeiro 2 1 State University of Rio de Janeiro, UERJ Civil Engineering Post-Graduate Programme, PGECIV e-mail: genasil@yahoo.com.br 2 State University of Rio de Janeiro, UERJ Mechanical Engineering Department, MECAN e-mail: jgss@uerj.br, soeiro@uerj.br Keywords: structural damage assessment, system identification techniques, inverse problems, differential evolution optimization technique, ant colony optimization technique. Abstract. Structural systems in a variety of applications including aerospace vehicles, automobiles and engineering structures such as tall buildings, bridges and offshore platforms, accumulate damage during their service life. The approach used in this investigation is one where the structural properties of the analytical model are varied to minimize the difference between the analytically predicted and empirically measured response. This is an inverse problem where the structural parameters are identified. In this work a reduced number of vibration modes and nodal displacements were used as the measured response. For the damage assessment problem a finite element model of the structural system is available and the model of the damaged structure will be identified. Damage will be represented by a reduction in the elastic stiffness properties of the system. In this investigation, the Differential Evolution (DE) and the Ant Colony Optimization (ACO) were applied to simple truss structures with different levels of damage. 1 INTRODUCTION In a typical load bearing structure, degradation of structural properties due to damage manifests itself as a change in the static and dynamic response. A correlation of the measured response with that obtained from an analytical model of the undamaged structure, allows for the possibility of determining a modified model that predicts the altered response. This inverse problem is solved using a system identification technique [1]. In this paper the output error approach of system identification is used to determine changes in the structural parameters that result from structural damage. Damage is represented by reduction in the elastic properties of the element. The net changes in these quantities due to damage are lumped into a single coefficient d i for each element that is used to multiply the stiffness matrix of that particular element. These coefficients d i constitute the design variables for the resulting optimization problem. Static displacements and eigenmodes are used as measured data for the inverse problem of damage detection. Reduced sets of eigenmodes and static displacements are used [2-3]. The approach of considering one design variable d i for each element in the structure usually results in a large dimensionality problem. These results in a very nonconvex design space, probably with several local minima, where Gradient- based nonlinear methods for function minimization may have difficulties to find the global optimum. In this work two global optimization methods, the Differential Evolution and the Ant Colony Optimization which are two heuristic population based methods were used for function minimization.