摘要
传统利用灰色关联分析方法对地震波动强度变化进行数学建模分析与仿真时,对地震波动强度变化的数列进行仿真分析时,忽略了地震波动强度的时间属性对结果的影响,导致分析结果准确性较低。本论述提出新的地震波动强度变化数学建模分析与仿真方法,通过地震波动强度序列的经验分布确定门限自回归模型的门限值,依据该门限值、AIC最小准则以及最小残差平方等方法获取地震波动强度序列的门限自回归模型,分析自回归模型的极限环和振荡的属性特点,得到地震波动强度变化的初步数值模拟结果。本论述构建了基于均生函数的地震波动强度序列的数学模型,通过均生函数数学建模方法拟合地震波动强度时间序列,依据时间序列基于双评分准则选取拟合周期,实现地震波动强度的数值仿真。实验结果表明,所提方法对地震波动强度变化模型具有较高的准确性和稳定性。
When the traditional grey prediction method is used to model and simulate variation in seismic wave intensity,the series of the variation is predicted without analyzing the time attribute of earthquake wave intensity.Hence,the accuracy of the analysis result is low.A new mathematical modeling analysis and simulation method for these variations is proposed in this paper.The threshold value of the threshold autoregressive model is determined by the empirical distribution of the seismic wave intensity series.The threshold autoregressive model of seismic wave intensity series is obtained by using the threshold value of the minimum criterion AIC and the least residual square method.The characteristics of limit cycles and oscillations of the auto-regressive model are analyzed,and the seismic wave intensity is obtained.The mathematical model of seismic wave intensity series is constructed based on the mean generating function.The time series of the earth-quake wave intensity is fitted by using the mean generation function.The fitting period is selected based on the double score criterion according to the time series.An accurate numerical simulation of seismic wave intensity is realized.The experimental results show that the proposed method is accurate and stable for the seismic wave intensity variation model.
作者
柴瑞帅
CHAI Ruishuai(Henan Institute of Economics and Trade, Zhengzhou 450046, Henan, Chin)
出处
《地震工程学报》
CSCD
北大核心
2018年第3期549-554,共6页
China Earthquake Engineering Journal
基金
河南省软科学课题(172400410212)
河南省重大科技专项项目(162102310147)
关键词
地震波动强度
数学建模
地震波动强度序列
均生函数
双评分准则
门限自回归模型
seismic wave intensity
mathematical modeling
seismic wave intensity sequence
mean generating function
couple score criterion
threshold autoregressive model
