Predicting geotechnical parameters from seismic wave velocity using artificial neural networks

Authors: JOHORA, FATEMA - University Of Mississippi; HICKEY, CRAIG - University Of Mississippi; YASARER, HAKAN - University Of Mississippi

Submitted to: Applied Sciences
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/12/2022
Publication Date: 12/13/2022
Citation: Johora, F.T., Hickey, C.J., Yasarer, H. 2022. Predicting geotechnical parameters from seismic wave velocity using artificial neural networks. Applied Sciences. 12:12815. https://doi.org/10.3390/app122412815.
DOI: https://doi.org/10.3390/app122412815

Interpretive Summary: Geotechnical investigation plays an indispensable role in site characterization and provides necessary data for various construction projects. However, geotechnical measurements are time consuming, point-based, and invasive. In this work, artificial neural networks models are developed to forecast geotechnical parameters from seismic wave velocity. Specifically, published seismic wave velocity, liquid limit, plastic limit, water content, and dry density from field and laboratory measurements are used to develop artificial neural networks models. Results showed that seismic wave velocity helps to predict water content and dry density. Conclusions reveal non-destructive geophysical measurements (seismic wave velocity) can complement geotechnical measurements to save project money and time. Geotechnical investigations conducted in this study are useful to those interested in obtaining data about the subsurface soil and rock conditions of a proposed development site. The results of this study help to understand the foundation requirements for the construction of any new infrastructures, underground utilities, underground parking lot and surrounding parking areas.

Technical Abstract: Geotechnical investigation plays an indispensable role in site characterization and provides necessary data for various construction projects. However, geotechnical measurements are time consuming, point-based, and invasive. Non-destructive geophysical measurements (seismic wave velocity) can complement geotechnical measurements to save project money and time. However, correlations between geotechnical and seismic wave velocity are crucial in order to maximize the benefit of geophysical information. In this work, artificial neural networks models are developed to forecast geotechnical parameters from seismic wave velocity. Specifically, published seismic wave velocity, liquid limit, plastic limit, water content, and dry density from field and laboratory measurements are used to develop artificial neural networks models. Due to the small number of data, models are developed with and without the validation step in order to use more data for training. The results indicate that the performance of the models is improved by using more data for training. For example, predicting seismic wave velocity using more data for training improves the R-squared value from 0.50 to 0.78 and reduces the average squared error from 0.0174 to 0.0075, and mean absolute relative error from 30.75 to 18.53. The benefit of adding velocity as an input parameter for predicting water content and dry density is assessed by comparing models with and without velocity. Models incorporating the velocity information show better predictability in most cases. For example, predicting water content using field data including the velocity improves the R-squared from 0.75 to 0.85 and reduces the average squared error from 0.0087 to 0.0051, and mean absolute relative error from 10.68 to 7.78. A comparison indicates that artificial neural networks models outperformed multilinear regression models. For example, predicting seismic wave velocity using field plus lab data has an artificial neural networks derived R2 value that is 81.39% higher than regression model.