Clough-Penzien功率谱模型参数的识别与统计建模及应用  

Identification and statistical modeling with applications of Clough-Penzien power spectrum model parameters

作　　者：丁佳伟 吕大刚[1,2] 曹正罡[1,2] DING Jia-wei;LÜDa-gang;CAO Zheng-gang(Key Lab of Structures Dynamic Behavior and Control of the Ministry of Education,Harbin Institute of Technology,Harbin 150090,China;Key Lab of Smart Prevention and Mitigation of Civil Engineering Disaster of the Ministry of Industry and Information Technology,Harbin Institute of Technology,Harbin 150090,China)

机构地区：[1]哈尔滨工业大学结构工程灾变与控制教育部重点实验室,黑龙江哈尔滨150090 [2]哈尔滨工业大学土木工程智能防灾减灾工业和信息化部重点实验室,黑龙江哈尔滨150090

出　　处：《振动工程学报》2023年第5期1204-1215,共12页Journal of Vibration Engineering

基　　金：国家重点研发计划资助项目(2021YFB2600500);国家自然科学基金资助项目(52078176,51678209)。

摘　　要：地震动的Clough-Penzien功率谱(简称“C-P谱”)模型具有明确的物理意义,但是该模型需要识别的参数较多;而普通最小二乘(OLS)算法存在对数据的非线性识别初始值选择要求高、无法快速有效实现对大批量参数的精确识别等缺点。本文采用自适应加权粒子群优化(AWPSO)算法,基于典型的地震记录对C-P谱模型参数进行识别。结果表明:AWPSO算法识别的精度相比于OLS算法至少提升2.3%,且在计算效率方面具有较大的提升。从地震记录数据库中挑选出4159条地震记录,按照建筑抗震设计规范中的场地分类标准对其分组,采用AWPSO算法对各类场地的C-P谱模型参数进行识别和统计。采用K-S检验、A-D检验及AIC准则,确定了各参数的最优概率分布模型,并进一步计算了各参数之间的相关系数,建立了C-P谱模型的联合概率密度函数。采用Latin超立方抽样方法,得到了场地的统计抽样C-P谱模型,与规范转换的功率谱进行了对比分析,利用功率谱迭代修正的人工地震动合成方法,生成了具有场地特性的人工地震记录。The Clough-Penzien power spectrum of ground motion(C-P spectrum)has clear physical significance,but needs to identify many parameters.The ordinary least squares(OLS)algorithm has shortcomings such as high requirements for the selec-tion of initial values for nonlinear identification of data,and inability to quickly and effectively realize accurate identification of large-scale parameters.In this paper,the adaptive weighted particle swarm optimization(AWPSO)algorithm is used to identify the C-P spectral model parameters based on typical ground motion records.The results show that the recognition accuracy of the AWPSO algorithm is improved by at least 2.3%compared with the OLS algorithm,and the computational efficiency has been greatly im-proved.4159 ground motion records are selected from the ground motion record database and grouped according to the site classifi-cation standards in the building seismic design code.The AWPSO algorithm is used to identify and count the parameters of the C-P spectrum model of various sites.Using K-S test,A-D test and AIC information criterion,the optimal probability distribution mod-el of each parameter is determined,and the correlation coefficient between each parameter is further calculated,and the joint proba-bility density function of C-P spectral model is established.Using the Latin hypercube sampling method,the statistical sampling C-P spectral model of the site is obtained,and the power spectrum of the standard transformation is compared and analyzed.The arti-ficial ground motion synthesis method with iterative correction of the power spectrum is used to generate the artificial seismic record with the site characteristics.

关 键 词：地震动合成 参数识别 统计建模 Clough-Penzien功率谱 AWPSO算法 

分 类 号：P315.9[天文地球—地震学]