Early Warning Classification of Geotechnical Engineering Construction Safety using Clique Graph Convolutional Network with Fast Sigmoid Activation

Through the fast growth of infrastructure development, the geotechnical engineering construction safety measure becomes the significant attention because of its complexity and the variation. Hastening the data of the geotechnical engineering construction safety measure will provides to the project management. However, the data involved in the geotechnical engineering are does not effectively combined due to there is no unified data standards. Hence, this research proposes the Clique Graph Convolutional Network with the Fast Sigmoid Activation (CGCN-FSA) function approach for the early warning classification of the Geotechnical Engineering construction safety measure. The FSA function delivers the smooth and non-linear transformation of the data, which supports the CGCN to model the complex patterns and relationships into geotechnical engineering data. The Geographic Information System (GIS) provides the significant understandings into standard surface situations, while Building Information Modelling (BIM) offers the detailed data about assets. The experimental results demonstrates that the proposed CGCN-FSA approach attains the better estimation time of 198.32s respectively as compared to the existing approach like GCN and Graph Neural Network (GNN).