In order to manage the safety of ground structures, it is important to accurately estimate the spatial distribution of geotechnical properties. The estimation method using Gaussian process regression requires the setting of appropriate random field parameters. Since the number and interval of data are important when estimating the parameters, we evaluate the accuracy of the estimated parameters using 1000 simulated data and compare them with the accuracy estimated by the Hessian matrix. It is confirmed that the Hessian matrix is useful for evaluating the parameters’ accuracy. The calculation cost of estimating the parameters of the random field by the maximum likelihood method becomes an obstacle when the number of data is large. It is indicated that the calculation cost is greatly reduced by introducing the Kronecker product. Depth-dependent non-stationarity is considered for the standard deviation of the random field, which is generally assumed to be stationary in many existing studies. Through the accurate evaluation of the estimated value by the blind test at several points on the measured data, the estimated accuracy with depthdependent non-stationarity is a more rational estimate compared with the case with the stationarity assumption.

Anomaly detection for infrastructure by deep learning using supervised learning has problems that it is difficult to prepare data set on abnormal data, consequently unlearned deterioration cannot be detected. On the other hand, several prior works have shown the effective for the anomaly detection based on generative models training using only normal data in image recognition tasks with little training data and diversity of the deteriorations to be detected. In particular, VAE-based methods have been used for anomaly detection on uniform parts. For the purpose of building the anomaly detection method of revetment that is independent of the type of deterioration, we examined and considered for the anomaly detection of revetment in block units using VAE. This paper empirically clarifies the guideline for preparation of data set and proposes an anomaly calculation method suitable for revetment. We also discussed t detecting various types of deteriorations.

The temporary pier structures are designed to withstand earthquakes in consideration of its short service life, and seismic input exceeding the design is also anticipated. In order to propose a temporary pier structure with a fuse mechanism that acts in the event of excessive seismic input, elemental tests and FEM analysis were conducted on the mortar fuse material used for the fuse mechanism to confirm the fracture behavior and to study the evaluation method of bearing capacity. The results showed that compressive failure with the top of the mortar fuse as the failure surface triggers the failure, and that it is possible to evaluate the bearing capacity of the mortar fuse by the mortar strength. Based on these results, static cyclic loading tests and dynamic loading tests were conducted using a model specimen to confirm the deformation behavior of the fuse mechanism, and the expected load-carrying capacity and dynamic characteristics were shown, confirming the validity of the fuse.

This study conducted a numerical analysis in order to clarfy the effect of train/bridge dynamic interaction on the dynamic response of steel and composite bridge with small masses. The result revealed that the resonance velocity decreased with decreasing bridge/train mass ratio and the impact factor at resonance decreased with decreasing stiffness and bridge/train mass ratio; however, the effects of stiffness and the mass ratio of bridge were small when the resonance phenomenon did not occur. The dynamic response of vehicles increases due to the deflection shape and resonance of the bridge. The effect of resonance decreases as the span of the bridge increases. Furthermore, for a mass ratio equivalent to an open-deck steel girder, the dynamic interaction effects reduced the dominant frequency of the bridge by 10% and the impact factor at resonance by 90%.