Abstract
A categorization of subsoil characterization approaches is used as the basis for evaluating them according to their capacity for model updating, quantifying uncertainties and modelling complex geometries. The information of using the right model at the right time is one of the most important sources of engineering knowledge. To ensure an equitable comparison under comparable starting conditions, three synthetic soil topologies are generated, each featuring relevant geological processes for subsoil characterization. In this study, we evaluated the performance of a voxel-based approach for subsoil characterization using the discrete output of Gaussian Process Regression (GPR). The evaluation based on the three metrics shows that the potential advantages are simplicity, ability to integrate new data and quantification of prediction uncertainty. Future work will focus on analysing additional approaches, such that an appropriate framework for digital-twins for geotechnical design and assessment can be defined.
Original language | English |
---|---|
State | Published - 2023 |
Event | 30th International Conference on Intelligent Computing in Engineering 2023, EG-ICE 2023 - London, United Kingdom Duration: 4 Jul 2023 → 7 Jul 2023 |
Conference
Conference | 30th International Conference on Intelligent Computing in Engineering 2023, EG-ICE 2023 |
---|---|
Country/Territory | United Kingdom |
City | London |
Period | 4/07/23 → 7/07/23 |