东南大学学报（自然科学版）2025，Vol.55Issue(6)：1572-1581,10.DOI:10.3969/j.issn.1001-0505.2025.06.009
考虑地形效应的地震动强度参数预测
Ground motion intensity parameter prediction considering topographic effects
摘要
Abstract
To further analyze the influence of topographic effects on ground motion intensity,a Bayesian neu-ral network model with complex topographic parameters was proposed.Strong ground motion records in Cali-fornia,USA,and terrain data around the sites were obtained based on the NGA-West2 strong motion database and global digital elevation model.A ground motion prediction model was established using the Bayesian neu-ral network method.The proposed model was compared with the Campbell-Bozorgnia model to analyze the in-fluence of topographic effects on ground motion intensity under ridge and valley conditions.The results show that topographic parameters play an important role in ground motion prediction.The proposed Bayesian neural network model with complex terrain parameters shows excellent prediction performance.It effectively cap-tures the variation of ground motion intensity under different terrain conditions,enhancing the accuracy of ground motion intensity prediction.Compared with the Campbell-Bozorgnia model,the prediction accuracy of the propoesd model increases by up to 12.10%under hilltop conditions and 15.09%under valley conditions.关键词
地震动预测方程/地形效应/贝叶斯神经网络/地震动强度预测Key words
ground motion prediction equation/topographic effect/Bayesian neural network(BNN)/ground motion intensity prediction分类
天文与地球科学引用本文复制引用
柯珂,刘坚,游田,徐龙河,谢行思..考虑地形效应的地震动强度参数预测[J].东南大学学报（自然科学版）,2025,55(6):1572-1581,10.基金项目
国家重点研发计划资助项目(2023YFC3805002) （2023YFC3805002）
国家自然科学基金资助项目(52322802). （52322802）
