Optimization Analysis Of Parameter Model Of Metro Shield Tunneling

Subway,a main system of urban rail transit,has become the key of public infrastructure construction in major cities both at home and abroad.Thanks to its safety,speed and low impact on ground traffic,the shield method has become the main way for subway tunnel construction.Because the tunneling of the shield machine is mainly realized by controlling various tunneling parameters,studying the characteristics of the tunneling parameters of the shield machine and establishing a reasonable tunneling parameter model are of great significance to ensure the safe and effective tunneling of the shield machine.This article is based on the interval between Tianjian Garden Station and Longcheng Middle Road Station of Shenzhen Metro Line 16.Perform data preprocessing based on the actual construction characteristics,and divide the stratum group of the left and right shield sections.Mathematical statistics and visualization methods are used to analyze the features of the main excavation parameters of each layer.And use multiple linear regression and random forest to perform For the prediction of tunneling parameter values,the association algorithm is used to mine the matching of tunneling parameter values.The main contents are as follows:(1)During the preprocessing the original data of shield tunneling parameters,a preprocessing method based on the actual construction characteristics of the shield is proposed.After preprocessing,the data can more conveniently and effectively reflect the tunneling situation of the shield,which is conducive to data description and models.The analysis proceeded smoothly;in the process of shield driving,the ground uplift monitoring data was analyzed,and the requirements of the specification were met,indicating the partial safety of the tunneling parameter setting.(2)Data description and correlation analysis are carried out on the main tunneling parameters of the overall tunneling processing of the left and right line shields.According to the results,the main tunneling parameters are roughly in the same variation,and the correlation coefficients between tunneling parameters are mostly similar.(3)Divided the strata group section of the left and right line shields through the strata,and described the main tunneling parameters under various strata conditions.The results showed that the shield tunneling parameters totally have different features under various strata conditions.(4)Predict the main excavation parameter values under various stratum conditions.Multiple linear regression method and random forest method are used to visually analyze the model error,and three indicators are used for quantitative evaluation.All of the results tells that the random forest method has higher prediction accuracy,and the it is improved in the range of 16% to 81%,which has reached the optimization of the prediction model method.(5)Perform association rule mining on the matching of the main excavation parameter values under different stratum conditions.First,the main excavation parameters are divided into five intervals at equal distances,then field coding is performed,and finally use strong association rules to mine to establish a parameter matching model.