A Quantitative Seismic Topographic Effect Prediction Method Based upon BP Neural Network Algorithm and FEM Simulation

• Published in: Journal of earth science (Wuhan, China) Vol. 35; no. 4; pp. 1355 - 1366
• Main Authors: Jiang, Qifeng, Rong, Mianshui, Wei, Wei, Chen, Tingting
• Format: Journal Article
• Language: English
• Published: Wuhan China University of Geosciences 01.08.2024; Springer Nature B.V; Shandong Institute of Earthquake Engineering,Jinan 250021,China%Key Laboratory of Urban Security and Disaster Engineering of China Ministry of Education,Beijing University of Technology,Beijing 100124,China%Publicity and Education Center,Shandong Earthquake Agency,Jinan 250014,China%Shandong Earthquake Station,Shandong Earthquake Agency,Jinan 250014,China; Shandong Seismic Hazard Prevention Center,Shandong Earthquake Agency,Jinan 250014,China
• Subjects: Algorithms; Amplification; Back propagation networks; Biogeosciences; Computer simulation; Earth and Environmental Science; Earth Sciences; Finite element method; Geochemistry; Geology; Geotechnical Engineering & Applied Earth Sciences; Ground motion; Mathematical models; Neural networks; Performance prediction; Predictions; Response spectra; Seismic response; Seismograms; Spectra; Topographic effects; Topography; earthquake disaster prevention; finite element method; seismic topographic effect; BP neural network algorithm; earth-quake disaster prevention
• ISSN: 1674-487X; 1867-111X
• DOI: 10.1007/s12583-022-1795-x

Topography can strongly affect ground motion, and studies of the quantification of hill surfaces’ topographic effect are relatively rare. In this paper, a new quantitative seismic topographic effect prediction method based upon the BP neural network algorithm and three-dimensional finite element method (FEM) was developed. The FEM simulation results were compared with seismic records and the results show that the PGA and response spectra have a tendency to increase with increasing elevation, but the correlation between PGA amplification factors and slope is not obvious for low hills. New BP neural network models were established for the prediction of amplification factors of PGA and response spectra. Two kinds of input variables’ combinations which are convenient to achieve are proposed in this paper for the prediction of amplification factors of PGA and response spectra, respectively. The absolute values of prediction errors can be mostly within 0.1 for PGA amplification factors, and they can be mostly within 0.2 for response spectra’s amplification factors. One input variables’ combination can achieve better prediction performance while the other one has better expandability of the predictive region. Particularly, the BP models only employ one hidden layer with about a hundred nodes, which makes it efficient for training.