MIC-based Correlation Analysis Between Ground And Building Settlement In Metro Construction

With the acceleration of urbanization in China, a large number of people poured intothe cities, and also there were frequent exchanging of goods. Thus, urban transport isfacing a severe situation, and the rapid development of urban subways is happening inchina. There are many uncertainties in the subway constructions period because they arevery complex in geological conditions, large in project scale, long in construction period,and also the project itself is underground engineering. Therefore, there are many potentialrisks in construction projects. Settlement is an important aspect of safety risk monitoring,because the settlement is an important indicator of the stability of the ground structure.Scholars have done a lot of researches about the settlement, mainly focused on qualitativeresearch of influence factors and subsidence prediction. The size and scope of theinfluence induced by metro construction still cannot be quantitatively described.Firstly, the monitoring objectives of ground settlement and building settlement isdescribed. The supervised accuracy of the settlement, monitoring frequency and themonitoring period is presented. According to the characteristics of monitoring of groundand building settlement in subway construction, settlement data preprocessing program ischosen. By comparing with Pearson correlation coefficient, Mutual information andMaximum information coefficient, the Maximum information coefficient (MIC) method isselected for the quantitative correlation analysis of ground subsidence and buildingsettlement.Under the background of Wuhan Jianghan Road Station subway construction project,the building settlement and ground settlement is adopted. Use the fitting models todescribe the trend of correlation coefficients and corresponding distance between theground settlement and building settlement. Through statistical analysis, the characteristicsof correlation between ground settlement and building settlement in each distance sectionare summarized, in order to determine the distance that the building would be impacted bymetro construction. Thus, buildings in these areas with strong influence should besupervised carefully. Though this way, advices could be provided for the risk analysis ofmetro construction.