Clogging Risk Early Warning for Slurry Shield Tunneling in Mixed Mudstone–Gravel Ground: A Real-Time Self-Updating Machine Learning Approach

Clogging constitutes a significant obstacle to shield tunneling in mudstone soils. Previous research has focused on investigating the influence of soils and slurry properties clogging, although little attention been paid impact parameters particularly early clogging warning during tunneling. This paper contributes developing real-time early-warning approach, based self-updating machine learning method. The judgment criteria are statistical characteristics whole-ring parameters. proposes use random forest (RF) for strategy clogging. performance this approach is illustrated through its application slurry-pressure-balanced construction Nanning metro line 1. Results show that RF-based can predict ring with only four minutes data, an accuracy 95%. RF model provided best compared other methods. Furthermore, realize accurate prediction one ring, using less data mechanism.