A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Artificial Neural Network–Based Inversion for Leaky Rayleigh Wave Dispersion Curve from Non-contact SASW Testing of Multi-layer Pavements
https://orcid.org/http://orcid.org/0000-0001-6300-3303
The air-coupled spectral analysis of surface waves (SASW) is an improvement technique of the conventional SASW method. The test can be used to evaluate the elastic modulus profile and layer thickness in the layered structures, such as pavement and soil ground. In air-coupled SASW test, the ground-coupled sensors are replaced by non-contact sensors to collect leaky surface waves, instead of the ground vibration. The evaluation is based on leaky Rayleigh wave dispersion characteristic in a layered media and then through a process of inversion which is considered a complicated process. This paper describes a development of an automated inversion procedure of non-contact SASW test data based on the neural network method. The effect of the structural properties of the pavement profile on the dispersion curves’ shapes was investigated too. The artificial neural network (ANN) models were trained by using a synthetic dispersion curves calculated from numerical simulation of different asphalt concrete pavement layer system configurations. The effect of the most optimized networks with the lower error rate and iteration number for convergence were selected and tested for certainty. The final results for the developed ANN models were reasonable and very close to the actual output.
Artificial Neural Network–Based Inversion for Leaky Rayleigh Wave Dispersion Curve from Non-contact SASW Testing of Multi-layer Pavements
https://orcid.org/http://orcid.org/0000-0001-6300-3303
The air-coupled spectral analysis of surface waves (SASW) is an improvement technique of the conventional SASW method. The test can be used to evaluate the elastic modulus profile and layer thickness in the layered structures, such as pavement and soil ground. In air-coupled SASW test, the ground-coupled sensors are replaced by non-contact sensors to collect leaky surface waves, instead of the ground vibration. The evaluation is based on leaky Rayleigh wave dispersion characteristic in a layered media and then through a process of inversion which is considered a complicated process. This paper describes a development of an automated inversion procedure of non-contact SASW test data based on the neural network method. The effect of the structural properties of the pavement profile on the dispersion curves’ shapes was investigated too. The artificial neural network (ANN) models were trained by using a synthetic dispersion curves calculated from numerical simulation of different asphalt concrete pavement layer system configurations. The effect of the most optimized networks with the lower error rate and iteration number for convergence were selected and tested for certainty. The final results for the developed ANN models were reasonable and very close to the actual output.
Artificial Neural Network–Based Inversion for Leaky Rayleigh Wave Dispersion Curve from Non-contact SASW Testing of Multi-layer Pavements
https://orcid.org/http://orcid.org/0000-0001-6300-3303
The air-coupled spectral analysis of surface waves (SASW) is an improvement technique of the conventional SASW method. The test can be used to evaluate the elastic modulus profile and layer thickness in the layered structures, such as pavement and soil ground. In air-coupled SASW test, the ground-coupled sensors are replaced by non-contact sensors to collect leaky surface waves, instead of the ground vibration. The evaluation is based on leaky Rayleigh wave dispersion characteristic in a layered media and then through a process of inversion which is considered a complicated process. This paper describes a development of an automated inversion procedure of non-contact SASW test data based on the neural network method. The effect of the structural properties of the pavement profile on the dispersion curves’ shapes was investigated too. The artificial neural network (ANN) models were trained by using a synthetic dispersion curves calculated from numerical simulation of different asphalt concrete pavement layer system configurations. The effect of the most optimized networks with the lower error rate and iteration number for convergence were selected and tested for certainty. The final results for the developed ANN models were reasonable and very close to the actual output.
Artificial Neural Network–Based Inversion for Leaky Rayleigh Wave Dispersion Curve from Non-contact SASW Testing of Multi-layer Pavements
Transp. Infrastruct. Geotech.
Al-Adhami, Hiba (author) / Gucunski, Nenad (author)
Transportation Infrastructure Geotechnology ; 8 ; 1-11
2021-03-01
11 pages
Article (Journal)
Electronic Resource
English
Implementation of Artificial Neural Networks to Automate SASW Inversion
| British Library Conference Proceedings | 2006
Effects of obstacles on Rayleigh wave dispersion obtained from the SASW test
| Online Contents | 1996
Effects of obstacles on Rayleigh wave dispersion obtained from the SASW test
| British Library Online Contents | 1996