Impact localization on a composite plate based on error outliers with Pugh's concept selection

Many researchers have performed impact localization using multiplexed fiber Bragg grating (FBG) sensors for a unique estimation of impact position. However, the previous impact localization did not show the methodology to choose the location with the lowest error among candidate impact points, estimated from sensors. We propose an error-outlier-based algorithm with Pugh's concept selection that can choose the impact point with the lowest error among several candidate impact points. The proposed impact localization algorithm was implemented for 30 arbitrary points on a carbon-fiber-reinforced polymer composite plate with two bonded FBG sensors. The results showed that the proposed algorithm could efficiently localize all impact points. Additionally, it could successfully choose the impact point with lower error among two candidate impact points estimated from the two sensors. The accuracy of selection was 93.33% and the totally estimated averaged-error for the selected sensors by Pugh's concept selection for all tests was 6.24 mm. Therefore, we conclude that the proposed impact localization algorithm is effective in localizing the impact points with the lowest errors when multiplexed sensors are used, regardless of the type of sensor. This technique will contribute to reducing measurement errors when localizing impact points on composite structures.