| 基于时序模型的变形预报研究 |
| |
| 引用本文： | 谢瑞,刘小强,李勇.基于时序模型的变形预报研究[J].测绘科学,2012(4):96-98. |
| |
| 作者姓名： | 谢瑞 刘小强 李勇 |
| |
| 作者单位： | 河南工程学院土木工程系;同济大学土木工程学院测量系;青岛市勘察测绘研究院 |
| |
| 基金项目： | 河南省教育厅自然科学研究计划(2011B420002) |
| |
| 摘 要： | 时序模型是一种有效的变形预报方法,但在很多变形预报时序模型应用中,存在建模过程不严密、缺乏统计检验和模型验证等问题。针对这些问题,本文首次在变形预报中采用LM检验来判断序列的相关性,ADF检验判断序列的平稳性,自相关图和偏自相关图初步确定模型类型和阶数以及AIC准则来确定最终模型,确保建模的每一过程都有严格的统计意义。文中利用实测沉降数据,建立了预报模型,并利用模型进行预报;最后将预报结果与实测数据比较,验证了预报模型的有效性,证明本文建模方法科学合理,能用于实际变形预报。
|
| 关 键 词： | 时间序列分析 LM检验 ADF检验 自相关函数 平稳性 |
Model of time series analysis for deformation forecast |
| |
| XIE Rui,LIU Xiao-qiang,LI Yong.Model of time series analysis for deformation forecast[J].Science of Surveying and Mapping,2012(4):96-98. |
| |
| Authors: | XIE Rui LIU Xiao-qiang LI Yong |
| |
| Institution: | ③(①Department of Civil Engineering,Henan Institute of Engineering,Zhengzhou 451191,China;②Department of Surveying and Geo-informatics Engineering,Tongji University,Shanghai 200092,China;③Qingdao Research lnstitute of Survey and MappingShandong Qingdao 266071,China) |
| |
| Abstract: | Time-series model is an effective method for deformation forecast.But in much application of time-series model for deformation forecast,the modeling process is not strict,lacking of statistics test and model test.So for the first time the paper cited LM test to determine the correlation of sequence,ADF test to determine the stability of sequence,the autocorrelation figure and partial autocorrelation figure to determine the type and the order of model preliminarily,and AIC criterion to determine the final model in deformation forecast,to assure every modeling process have strict statistical significance.With a grou Pof deformation monitoring data,it established the forecast model and forecasts.Comparing forecast results with monitoring data,it verified the effectiveness of forecast model,could be used in actual deformation forecast. |
| |
| Keywords: | time series analysis LM test ADF test autocorrelation function stationarity |
| 本文献已被 CNKI 等数据库收录！ |
|
