Application of Multidimensional Association Mining in Shield Tunneling Optimization

Abstract

The rapid development of data science promotes the innovative application of data mining algorithms in geotechnical engineering. So the multidimensional association algorithm is tried in the technical optimization of shield tunneling. Based the original data sets on geological condition, tunneling operation and monitoring, the data preprocessing is carried out according to such steps as sorting and statistics, classification and coding, factor selection and data discretization. Then eight important parameters are selected for association analysis, including tunneling layer, surface settlement, and six operating parameters of the shield machine. Apriori algorithm is used for mining the association rules to establish the association models of those parameters, whose rationality is tested. Also the engineering significance of the multi-parameter association models is explained in the view of practical experience. Finally, the data mining principles of the shield tunneling parameters association rules are discussed. As a simple and reasonable data analysis method, the association algorithm provides a new way for the intelligent construction of shield tunneling.