首页 | 本学科首页   官方微博 | 高级检索  
     检索      

非平稳时序分析的B-J建模法在变形预测中的应用
引用本文:佘娣,谢劭峰,彭家頔,雷露露,杜红飞.非平稳时序分析的B-J建模法在变形预测中的应用[J].测绘科学,2012,37(5):207-208,226.
作者姓名:佘娣  谢劭峰  彭家頔  雷露露  杜红飞
作者单位:桂林理工大学测绘地理信息学院/广西空间信息与测绘重点实验室,广西桂林,541004
基金项目:广西自然科学基金项目资助(桂科基0991023)
摘    要:本文简要介绍了时序分析的基本原理,论述了非平稳序列的B-J建模法的步骤,包括数据预处理、模型识别、参数估计及预报等;分别用时序分析B-J法和回归分析法进行了实例建模及预报,并对预报结果进行了分析比较,论证了单纯的时序分析法短记忆预报的特点,讨论了时序分析的优缺点,最后总结了在进行变形预报时必须根据数据的数学特征选择相应的数学模型。

关 键 词:非平稳  时间序列分析  回归分析  预报

Application of B-J Model for non-stationary time-serial analysis in deformation prediction
SHE Di , XIE Shao-feng , PENG Jia-di , LEI Lu-lu , DU Hong-fei.Application of B-J Model for non-stationary time-serial analysis in deformation prediction[J].Science of Surveying and Mapping,2012,37(5):207-208,226.
Authors:SHE Di  XIE Shao-feng  PENG Jia-di  LEI Lu-lu  DU Hong-fei
Institution:(College of Geomatics and Geoinformation Guilin University of Technology/Guangxi Key Laboratory of Spatial Information and Geomatics,Guangxi Guilin 541004,China)
Abstract:The basic principles of time-serial analysis were introduced first and the steps of B-J method based on non-stationary series,including data preprocessing,model identification,parameter estimation and prediction,were discussed in this paper.B-J time-series analysis methods and regression analysis were used for modeling and forecasting,and the forecast results were analyzed and compared.It demonstrated the characteristics of a short memory forecasting of a simple time series analysis,and discussed the advantages and disadvantages of time series analysis.Finally it was concluded that during the deformation forecast the appropriate mathematical mode must be selected based on the mathematical characteristics of the data.
Keywords:non-stationary  time series analysis methods  regression analysis  prediction
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号