Key points are not available for this paper at this time.
ABSTRACT: This paper proposes a domain ontology for tunnel construction risk management. The construction process of the ontology is emphasized in this paper, encompassing top-level ontology design, trigger ontology, tunnel ontology, ground ontology, tunnel construction disaster terminology ontology, and the risk section. The ontology utilizes the OWL language to define properties, translating complex relationships in tunnel construction risk analysis into machine-readable language and supporting knowledge inference between entities. Furthermore, the paper introduces ontology inference rule design based on the SWRL rule language, enabling functionalities such as risk identification, assessment, and management through rule-based inference mechanisms. 1. INTRODUCTION 1.1. Background In practical risk management, personnel systematically assess on-site risk points using prepared risk control checklists and an indicator system. However, creating these checklists requires a high level of professionalism and on-site experience. The effectiveness of risk identification, comprehensiveness of assessment results, and accuracy of evaluation outcomes depend on the qualifications and experience of the checklist preparer. For instance, in addressing the risk of sudden water influx in underwater tunnels, experienced personnel consider factors like surrounding rock composition, adverse geology, and tunnel geometry. Transferring such implicit knowledge remains challenging, and the departure of experienced risk workers poses a significant loss for tunnel construction projects. Furthermore, the unique characteristics of each tunnel project complicate knowledge transfer in risk management across projects. Tunnel risk assessment is conducted in three stages: design, construction, and operation. At each stage, the tunnel risks are identified and assessed based on the geological survey report and relevant updated information. Underwater tunnel geological exploration is particularly challenging due to the limitations of technology, and the accuracy of exploration results needs improvement. Static assessment results compiled based on exploration reports cannot meet the demands of practical risk management. Therefore, it is essential to perform dynamic evaluations of tunnel construction risks based on continuously updated data during the construction process. There are two challenges that prevent dynamic risk management techniques from getting off the ground • On the one hand, dynamic risk assessment requires a comprehensive analysis of various data types, including geological survey data, advance geological forecasting data, monitoring measurement data, design plans, weather data, etc. However, these data belong to different units, each with its data storage platform, limited inter-unit data communication, and sometimes incompatible formats. • On the other hand, the data generated during tunneling has a temporal nature, and risk assessment through methods like numerical simulation is less efficient, making it challenging to respond dynamically to real-time updated data in complex decision scenarios.
Xue et al. (Sun,) studied this question.