Identification of three-dimensional defect topology in concrete structures based on self-attention network using hammering response data

抄録

The ageing of concrete structures in Japan is becoming an increasingly serious issue. Periodic inspection is necessary to prevent accidents caused by ageing. One of the methods used to inspect concrete is the hammering test. In this study, we aim to develop a system using machine learning for identifying the topology of defects in concrete, based on acceleration response data obtained from the hammering test. As part of the machine learning method, we constructed a neural network based on self-attention. In addition, we evaluated the effect of changing the size of the estimation domain using the machine learning model.