Displacement monitoring of large bridges is an important source of information concerning their health state. In this paper, a procedure based on satellite Persistent Scatterer Interferometry (PSI) data is presented to assess bridge health. The proposed approach periodically assesses the displacements of a bridge in order to detect abnormal displacements at any position of the bridge. To demonstrate its performances, the displacement characteristics of two bridges, the Nanjing-Dashengguan High-speed Railway Bridge (NDHRB, 1272 m long) and the Nanjing-Yangtze River Bridge (NYRB, 1576-m long), are studied. For this purpose, two independent Sentinel-1 SAR datasets were used, covering a two-year period with 75 and 66 images, respectively, providing very similar results. During the observed period, the two bridges underwent no actual displacements: thermal dilation displacements were dominant. For NDHRB, the total thermal dilation parameter from the PSI analysis was computed using the two different datasets; the difference of the two computations was 0.09 mm/°C, which, assuming a temperature variation of 30 °C, corresponds to a discrepancy of 2.7 mm over the total bridge length. From the total thermal dilation parameters, the coefficients of thermal expansion (CTE) were calculated, which were 11.26 × 10−6/°C and 11.19 × 10−6/°C, respectively. These values match the bridge metal properties. For NYRB, the estimated CTE was 10.46 × 10−6/°C, which also matches the bridge metal properties (11.26 × 10−6/°C). Based on a statistical analysis of the PSI topographic errors of NDHRB, pixels on the bridge deck were selected, and displacement models covering the entire NDHRB were established using the two track datasets; the model was validated on the six piers with an absolute mean error of 0.25 mm/°C. Finally, the health state of NDHRB was evaluated with four more images using the estimated models, and no abnormal displacements were found.

No abstract

With the continuous expansion of the high-speed railway network in China, long-span railway bridges carrying multiple tracks demand reliable and fast testing procedures and techniques. Bridge dynamic behavior analysis is a critical process in ensuring safe operation of structures. In this study, we present some experimental results of the vibration monitoring of a four-track high-speed railway bridge with a metro–track on each side: the Nanjing–Dashengguan high-speed railway bridge (NDHRB). The results were obtained using a terrestrial microwave radar interferometer named IBIS-S. The radar measurements were interpreted with the support of lidar point clouds. The results of the bridge dynamic response under different loading conditions, including high-speed trains, metro and wind were compared with the existing bridge structure health monitoring (SHM) system, underlining the high spatial (0.5 m) and temporal resolutions (50 Hz–200 Hz) of this technique for railway bridge dynamic monitoring. The detailed results can help engineers capturing the maximum train-induced bridge displacement. The bridge was also monitored by the radar from a lateral position with respect to the bridge longitudinal direction. This allowed us to have a more exhaustive description of the bridge dynamic behavior. The different effects induced by the passage of trains through different tracks and directions were distinguished. In addition, the space deformation map of the wide bridge deck under the eccentric load of trains, especially along the lateral direction (30 m), can help evaluating the running stability of high-speed trains.

Convolution-based autoencoder networks have yielded promising performances in exploiting spatial–contextual signatures for spectral unmixing. However, the extracted spectral and spatial features of some networks are aggregated, which makes it difficult to balance their effects on unmixing results. In this paper, we propose two gated autoencoder networks with the intention of adaptively controlling the contribution of spectral and spatial features in unmixing process. Gating mechanism is adopted in the networks to filter and regularize spatial features to construct an unmixing algorithm based on spectral information and supplemented by spatial information. In addition, abundance sparsity regularization and gating regularization are introduced to ensure the appropriate implementation. Experimental results validate the superiority of the proposed method to the state-of-the-art techniques in both synthetic and real-world scenes. This study confirms the effectiveness of gating mechanism in improving the accuracy and efficiency of utilizing spatial signatures for spectral unmixing.

No abstract