Damage identification based on topology optimization and Lasso regularization

In this paper, we present a damage identification method for small damages based on topology optimization and Lasso regularization. In particular, this work extends the applicability of the previously developed damage identification method using frequency response functions and topology optimization, by conducting rigorous parametric studies in terms of damping, measurement noise, and damage size. It is shown that the presented method successfully identifies small damaged regions with a reasonable accuracy. To evaluate the effectiveness of the proposed method, we applied the method to identify the damages in cantilevered plates that are subject to static or dynamic loads. The method succeeded in detecting the locations and shapes of damages more accurately than the method without Lasso regularization. Furthermore, in most cases we have considered, spurious damages generated during the optimization were successfully suppressed.