Improving Probability-Based Diagnostic Imaging Algorithm by Modeling Weight Distribution Coefficient for Instantaneous Baseline Damage Identification

This study used a combination of the instantaneous baseline method and a probability-based diagnostic imaging (PDI) algorithm for damage identification in a stiffener-reinforced aluminium plate. Selecting an appropriate weight distribution function plays a significant role in the PDI algorithm by expressing the probability of damage existence around a sensing path. Weight distribution functions are highly dependent on the weight distribution coefficient (β). This parameter (β) normally is estimated by trial and error, which is incorrect. In addition, determining this parameter is quite costly and time-demanding. To address these issues thoroughly, a new approach is proposed for modelling the weight distribution coefficient based on the physics of wave propagation and the actuator-sensor distance. An appropriate value for β is estimated for each sensing path between actuators and sensors in a specific structure based on the proposed model, making the results more accurate and reasonable. The proposed model was validated by identifying damages at three different locations of a structure for five scenarios (three single-damage and two multiple-damage scenarios). Moreover, the results obtained from proposed model using linear and Gaussian weight distributions were compared with the results extracted from an existing approach. The acceptable localization results clearly indicate the effectiveness of proposed approach.