A Two-Stage Approach to Solve Structural Damage Detection Problem in Plate Structures

Abstract

A two-stage optimization approach, based on modal strain energy and an Efficient Accelerated Particle Swarm Optimization algorithm (EAPSO), for localization and quantification of damage in plate-like structures is proposed. First, to localize damage sites, a new damage index is developed using modal strain energy and the statistical hypothesis analysis technique. Then, the proposed EAPSO algorithm is used to estimate the extent of damaged elements located in the first stage. The objective function to be minimized is established via flexibility Matrix changes of the structure. The EAPSO operates with a new updating model, created on random distribution of particles from the search space, which supports discovering and preserving interesting regions of the research space. Control parameters are fine-tuned to properly balance exploration/exploitation while accelerating convergence of EAPSO. In addition, a micro-search operator, which remove small damages from solutions, is embedded in the algorithm to decrease dimension of the search space and saves computational effort. To examine the effectiveness of the suggested approach, a numerical example for isotropic plate is carried out. The simulation results are compared with those of some recent algorithms, indicating the superiority of the proposed approach in terms of reliability, robustness to noise and rapidity of convergence.