Abstract: In order to make full use of a large number of monitoring data recorded by shield machine during tunnel construction, data analysis and mining techniques are adopted to preprocess the monitoring data. Then, with the help of association rule analysis method and improved Apriori algorithm, the relationship among more than ten parameters in shield construction is analyzed and mined, and the typical association results are explained, which verify the mastered laws, correct misunderstanding and discover the potential unknown laws. Finally, based on classification and regression tree algorithm, the realtime prediction of the stratum where the shield machine is located is made, and a threedimensional geological model is built to interpret and apply the prediction results. The results show that the classification model can reflect the stratum information more truly and can be used as an effective supplement to the geological survey results. The feasibility and rationality of analyzing and mining the monitoring data of shield machine is verified by Guangzhou metro section.

Key words: shield construction, monitoring data, data mining, association rules, Apriori algorithm, classification algorithm, classification and regression tree algorithm, soil classification