Intelligent Metro Shield Tunnel Structure Assessment Based on Knowledge Graph

Abstract

With urban metro being put into service, various structure defects, such as dislocation, leakage, crack, spalling, gradually happen. Traditionally, engineers or experts are required to inspect and assess the tunnel structure on site regularly. But most of the assessment are subjective and qualitative, failing to achieve reasonable assessment and accurate tunnel maintenance operation. Therefore, some quantitative methods, such as Tunnel Servility Index (TSI), are proposed. However, TSI method only assesses the current condition of a tunnel, without considering its historical monitoring data. With the development of geotechnical engineering, large amount of monitoring data has been recorded. These data are multi-source heterogeneous, making it rather difficult to store and analyse. Database technology, especially Knowledge Graph (KG), is good at storing, managing and mining multi-source heterogeneous data. Based on KG, all data related to tunnel structure can be stored and then taken into account when making assessment; besides, the information of a tunnel from its construction to the present can be recorded in the knowledge graph. In the paper, monitoring data of Shanghai Metro Line 1 is stored in Neo4j, a graph database, to form the metro tunnel knowledge graph. Based on the KG, the metro tunnel is assessed by a dynamic assessment model based on TSI method. The hided statistic relationships among defects are also deduced. Through the application of KG on Shanghai Metro Line 1, the dynamic assessment model is proposed to be effective, reliable and advanced.