Design of Tunnel Support System Based on Machine Learning

With the remarkable advancements in Artificial Intelligence (AI) and Machine Learning (ML) technologies across various fields, these tools offer innovative directions for long-standing, experience-based designs in tunnel and underground construction. Especially in the context of vulnerable rock tunnels under high geostress conditions, these technologies have opened up new avenues for research. In such settings, primary support structures frequently experience large-scale deformation due to rock squeezing, subsequently leading to structural failure. To address this issue, we introduce a novel passive deformation node based on the concept of energy release.Initially, we conducted laboratory tests to investigate the mechanical behavior of this new node. A series of rigorous tests were performed to understand its performance under various conditions. Following the lab experiments, field trials were carried out to validate the mechanical properties of this novel node.Ultimately, we employed the PSO-SVM (Particle Swarm Optimization-Support Vector Machine) algorithm to ascertain the data stability of the passive deformation node. The results were then integrated into the tunnel support design, providing further refinement to the support structures and construction plans