地铁地表沉降监测数据的预处理及预测方法的探讨

发布时间：2018-03-26 07:33

本文选题：变形监测　切入点：数据处理　出处：《东华理工大学》2015年硕士论文

【摘要】：随着我国新型城镇化建设的不断推进,大城市在资源和环境方面的承载力趋于饱和,不断增多的人口加重了交通拥挤和环境污染的程度,人们的生活质量严重下降,尤其是出行不便带来的困扰。大力发展城市地下轨道交通成了大城市解决问题的必由之路,这也给地铁变形监测领域带来了重大发展机遇,各种监测设备和技术相继问世,但是当前的技术仍无法代替人工测量。地铁变形监测要求数据具有准确性和及时性,与时间关联性很强,但是受地铁施工环境的影响,监测数据有时会出现差错、漏测等问题,这些问题深深困扰着一线的测量工作者们。本文的研究就是以解决这种情况为出发点,为一线的测量工作者提供多种解决问题的方法。变形监测的数据处理主要分为两大部分:数据的预处理阶段和数据的分析预测阶段。预处理阶段包括数据的粗差探测处理和缺失数据的插补处理,在介绍相应理论的同时,利用实测数据进行对比分析并对一些算法的缺陷做出改进,提高其准确性和可靠性。在数据分析建模阶段包括数据的拐点探测,时间序列的模型检验,建模过程等内容。根据数据的特点,采用了经过差分处理的ARIMA模型,在建模的过程中,大量采用图像和数据表格相结合的方式,使得建模过程更简便直观。本论文的实验数据来自青岛地铁二号线工程施工方变形监测实测的数据,其数据真实可靠,这对于相关领域的其他监测项目,有着一定的参考意义。
[Abstract]:With the development of new urbanization in China, the carrying capacity of large cities in resources and environment tends to saturation, the increasing population increases the degree of traffic congestion and environmental pollution, and the quality of life of people drops seriously. Especially the trouble caused by the inconvenience of travel. Vigorously developing urban underground rail transit has become the only way to solve the problem in large cities, which has also brought great development opportunities to the field of subway deformation monitoring, and various monitoring equipment and technologies have emerged one after another. However, the current technology can not replace manual measurement. Subway deformation monitoring requires that the data be accurate and timely, and have strong correlation with time. However, due to the influence of subway construction environment, the monitoring data will sometimes appear some problems, such as errors, missing measurements, etc. These problems are deeply perplexing the front-line surveyors. The purpose of this paper is to solve this problem. The data processing of deformation monitoring is divided into two parts: the data preprocessing stage and the data analysis and prediction stage. The preprocessing stage includes the gross error of the data. Detection processing and interpolation of missing data, At the same time of introducing the corresponding theory, we use the measured data to carry on the contrast analysis and make the improvement to some algorithm's defect, improve its accuracy and reliability. In the stage of the data analysis and modeling, it includes the data inflection point detection, the time series model checking, According to the characteristics of the data, the ARIMA model which is processed by difference is adopted. In the process of modeling, a large number of images and data tables are combined. The experimental data in this paper come from the actual data of deformation monitoring of the construction side of Qingdao Metro Line 2, and the data are true and reliable, which has certain reference significance for other monitoring projects in related fields.
【学位授予单位】：东华理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P642.26;U231.1

【参考文献】