Robust damage detection method against temperature variation using stochastic differential equations

In this paper, we introduce a new application of stochastic differential equations in damage detection for mechanical structures which is robust against variations in environmental conditions. Drift and diffusion functional forms of the model are obtained using a nonparametric estimator without any prior knowledge about the dynamic behavior of the system. After determining functional forms from instantaneous conditional means and variances, a parametric estimator is employed to estimate the value of equation parameters. Simulation results show that deterministic part has valuable information about the damage; make it a powerful tool for damage detection. An aluminum beam is considered as the case study for evaluating the performance of the proposed method. Also a comparison with a popular fault detection method has been presented.