马尔可夫链改进的MMF沉降预测模型及应用  

An improved MMF subsidence prediction model based on the Markov chain and its application

作　　者：赵亚红 王伟娜 江培华 李佳慧 ZHAO Yahong;WNAG Weina;JIANG Peihua;LI Jiahui(Architectural Engineering College,North China Institute of Science and Technology,Beijing 101601,China;School of Digital Construction,Shanghai Urban Construction Vocational College,Shanghai 200438,China)

机构地区：[1]华北科技学院建筑工程学院,北京101601 [2]上海城建职业学院数字建造学院,上海200438

出　　处：《测绘通报》2022年第1期79-83,共5页Bulletin of Surveying and Mapping

基　　金：河北省自然科学基金(E2019508162);廊坊市科技支撑计划(2020011038,2020011017,2021013088);河北省高等教育教学改革研究与实践项目(2019GJJG465)。

摘　　要：针对地基沉降机理复杂及随机性特点,结合马尔可夫链理论,本文建立了一种马尔可夫链改进的MMF沉降预测模型。首先采用部分实测沉降数据,利用CurveExpert软件拟合MMF模型;然后根据MMF模型预测相对误差大小,并按照马尔可夫理论划分状态区间,构建状态转移概率矩阵,预测下一个沉降量所处的状态,从而得到了马尔可夫链改进的MMF预测值;最后将本文模型应用于深圳滨海大道市政工程软土路基沉降预测中,并对模型的预测效果进行精度分析。结果表明,马尔可夫链改进的MMF模型的预测精度较单一的MMF有明显提高,建模方法合理,可用于类似的工程预测。Aiming at the complex and random characteristics of foundation settlement mechanism,an improved MMF settlement prediction model based on Markov chain theory is established.Firstly,the MMF model is fitted by CurveExpert software with some measured data.Then,according to the relative error predicted by the MMF model,the state interval is divided according to Markov theory,and the probability matrix of state transition is constructed to predict the state of the next settlement,thus the MMF prediction value improved based on Markov chain is obtained.Finally,the model is applied to the settlement prediction of shenzhen Binhai road municipal engineering soft soil subgrade,and the prediction accuracy of the model is analyzed.The results show that the prediction accuracy of the MMF model based on Markov chain is significantly higher than the MMF model.The modeling method is reasonable and can be used for similar engineering prediction.

关 键 词：马尔可夫链 MMF 地基 沉降预测 应用 

分 类 号：P258[天文地球—测绘科学与技术]