| EM算法的时序模型在沉降数据处理中的应用 |
| |
| 引用本文： | 马传宁,蔡伟,关沧海,徐琦.EM算法的时序模型在沉降数据处理中的应用[J].测绘科学,2017(12):178-184. |
| |
| 作者姓名： | 马传宁 蔡伟 关沧海 徐琦 |
| |
| 作者单位： | 国家地理信息局第二大地测量队(黑龙江第一测绘工程院),哈尔滨,150000 |
| |
| 基金项目： | 黑龙江省对地观测与导航工程技术创新平台项目 |
| |
| 摘 要： | 针对时间序列分析对监测数据中出现的不完全数据(部分数据缺失)无法进行精确建模的问题,该文引入期望极大算法(EM算法),提出EM算法与时间序列分析的组合算法模型。运用EM算法的时间序列分析组合算法模型可以对沉降过程中遇到的不完全沉降数据进行建模分析,该组合算法模型可以对不完全沉降数据进行较为精确的建模,并对后期沉降数据进行较为精确的预测。将某地铁基坑点沉降数据作为实验数据,EM算法的时间序列分析的建模结果表明:所提出的组合算法模型可以对不完全沉降数据进行建模分析,绝对误差为0.23mm,建模精度较高。
|
| 关 键 词： | 变形监测 EM算法 时间序列分析 组合算法模型 |
Application of time series model of EM algorithm in settlement data processing |
| |
| Abstract: | Aiming at the problem of incomplete modeling(partial data loss)in the monitoring data,the maximum algorithm(EM algorithm)was introduced,and a combined algorithm model of EM algorithm and time series analysis was proposed.The time series analysis and combinatorial algorithm model of EM algorithm can be used to model the incomplete settlement data encountered in the settlement process.The combined algorithm model could model the incomplete settlement data accurately and make the later settlement data more accurate prediction.The results of time series analysis of EM algorithm showed that the combined algorithm could model and analyze the incomplete settlement data,and the absolute error was 0.23 mm.The simulation results showed that the modeling precision of proposed method was high. |
| |
| Keywords: | deformation monitoring EM algorithm time series analysis combined algorithm model |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
