Tweet Share

• Send by email

  Send by email

  

  Tunnel Process Control for Klang Valley Metro Phase 2 Using Machine Learning and Artificial Intelligence

  The Klang Valley Mass Railway Transit in Kuala Lumpur is mid-way through line 2 construction. The line faces several challenges including mixed face conditions, variable ground water and geologic hazards with high potential for large settlements and sink holes. Consequently, the influences on tunnel processes such as advance rates and cutter tool wear are hard to estimate. This paper presents the use of several tools including geostatistics and machine learning/artificial intelligence to improve ground interpretation, predict machine performance and identify optimum management strategies using line 2 shield machine and geotechnical data. The outcome of the analysis provides a number of novel advances, including real time probabilistic assessment of exceeding various settlement levels along the alignment, the reduction in settlement probability as a function of shield parameters, and a data driven characterization of shield advance rate and tool wear that relates to controllable shield operation parameters. The analysis will ultimately provide more representative forecasting to evaluate the state of the project and aid in planning. Data sets produced through these techniques have value for planning and tendering for upcoming projects in Kuala Lumpur.

  Recipient :

  Name ^* :

  E-mail address ^* :

  ^* Required fields

  Send   or  Cancel

• Print