基于天牛须算法优化极限学习机的砂土地震液化预测  

Prediction of Sandy Soil Liquefaction Based on Extreme Learning Machine Optimized by the Beetle Antennae Search Algorithm

作　　者：吕国军 王晨晖[1,3] 王秀敏 袁颖 LV Cuojun;WANG Chenhui;WANG Xumin;YUAN Ying(National Field Scientific Observation and Research Station for Huge Thick Sediments and Seismic Disasters in Hongshan,Hebei Xingtai054000,China;Hebei Earthquake Agency,Hebei Shijiazhuang 050031,China;Xingtai Central Seismic Station,Hebei Earthquake Agency,Hebei Xingtai 054000,China;School of Urban Geology and Engineering,Hebei Geologic University,Hebei Shijiazhuang 050031,China;Hebei Province Underground Artificial Environment Intelligent Development and Control Technology Innovation Center,Hebei Shijazhuang 050031,China)

机构地区：[1]河北红山巨厚沉积与地震灾害国家野外科学观测研究站,河北邢台054000 [2]河北省地震局,河北石家庄050031 [3]河北省地震局邢台地震监测中心站,河北邢台054000 [4]河北地质大学城市地质与工程学院,河北石家庄050031 [5]河北省地下人工环境智慧开发与管控技术创新中心,河北石家庄050031

出　　处：《四川地震》2025年第2期1-6,共6页Earthquake Research in Sichuan

基　　金：国家自然科学基金(41807231);河北省重点研发计划项目(22375406D)。

摘　　要：为科学有效地预测砂土地震液化,采用唐山地区64组砂土地震液化实例,利用天牛须算法(Beetle Antennae Search Algorithm,BAS)和优化极限学习机(Extreme Learning Machine,ELM)的连接权值和阈值,建立基于BASELM的砂土地震液化预测模型。选取50个砂土液化实例训练模型的稳定性和有效性,并对剩余14个样本进行模型验证。结果表明:BAS-ELM模型的预测结果与实际值拟合度更高,其平均绝对误差、平均绝对百分误差及均方根误差分别为12%、5.667%和1.7%,均优于ELM模型,具有较好的预测精度。To scientifically and effectively predict seismic liquefaction of sandy soil,we used 64 cases of sand liquefaction in the Tangshan area to develop a sand liquefaction prediction model based on the Beetle Antennae Search Algorithm(BAS)optimized Extreme Learning Machine(ELM).Specifically,we employed the BAS algorithm to optimize the connection weights and thresholds of the ELM.We selected fifty cases of sand liquefaction to train the model and assess their stability and effectiveness,and used the remaining 14 cases for prediction.The results showed that the BAS-ELM model exhibited a higher degree of fit between the predicted and actual values.The mean absolute error,mean absolute percentage error,and root mean square error are 12%,5.667%,and 1.7%,respectively,which is better than the ELM model and shows superior prediction accuracy.

关 键 词：地震 砂土液化 天牛须算法 极限学习机 

分 类 号：P315[天文地球—地震学]