测绘学报 ›› 2017, Vol. 46 ›› Issue (4): 526-532.doi: 10.11947/j.AGCS.2017.20160531

• 工程测量 • 上一篇    下一篇

顾及不确定性影响的变形概率预报法

魏冠军1, 党亚民2, 章传银2, 杨维芳1   

  1. 1. 兰州交通大学测绘与地理信息学院, 甘肃 兰州 730070;
    2. 中国测绘科学研究院, 北京 100830
  • 收稿日期:2016-10-24 修回日期:2017-03-16 出版日期:2017-04-20 发布日期:2017-05-05
  • 通讯作者: 党亚民 E-mail:dangym@casm.ac.cn
  • 作者简介:魏冠军(1976-),男,博士,副教授,主要从事测量数据处理的理论与算法研究。
  • 基金资助:
    国家自然科学基金(41364001);甘肃省自然科学基金(1508RJEA065);兰州交通大学科技支撑计划(ZC2014002)

Method of Deformation Probability Prediction Considering the Influence of Uncertainty Factors

WEI Guanjun1, DANG Yamin2, ZHANG Chuanyin2, YANG Weifang1   

  1. 1. Faculty of Geomatics, Lanzhou Jiaotong University, Lanzhou 730070, China;
    2. Chinese Academy of Surveying and Mapping, Beijing 100830, China
  • Received:2016-10-24 Revised:2017-03-16 Online:2017-04-20 Published:2017-05-05

摘要: 针对变形预报的不确定性,以MCMC算法和贝叶斯预测理论为基础,提出了变形概率预报方法,该方法以概率分布的形式描述变形预报的不确定性,通过概率规则实现预报的递推过程。利用宁杭高速路基沉降数据进行数值计算,定量分析了预报值及其可靠性区间等信息,并与最小二乘估计、免疫算法的预报结果进行比较,结果表明了该方法的有效性和可行性。

关键词: 预报不确定性, 测量误差, 参数不确定性, Gibbs抽样, 概率预报

Abstract: A probabilistic prediction method of deformation is proposed based on the MCMC algorithm and the Bayesian Prediction Theory.This method describes the uncertainty of deformation prediction using probability distributions and implement the recursive process of prediction by probability rules. The settlement data from the Nanjing-Hangzhou high-speed roadbed is used to quantitatively analyze the forecast values, reliability intervals and so on, and then the results are compared with those obtained by the least squares estimation and the immune algorithm, and it has shown that the proposed method is effective and feasible.

Key words: prediction uncertainty, measurement error, parameter uncertainty, Gibbs sampling, probability predication

中图分类号: