Design of Tunnel Passive Deformation and Compression Support Based on SVM Machine Learning

With the rapid development of artificial intelligence, the adoption of machine learning methods has provided new insights for addressing the longstanding reliance on empirical design in tunnel and underground engineering. This study focuses on the problem of initial support structure damage caused by large deformations due to high ground stress weak surrounding rock in tunnels. Based on the concept of energy release, a novel passive deformation node is proposed. Through laboratory experiments on the node, its mechanical properties are analyzed. Subsequently, numerical simulations of the novel passive deformation node are conducted to validate its mechanical performance. Finally, the data reliability of the passive deformation node is verified using machine learning SVM (Support Vector Machine), and feedback is provided to the tunnel support system design for optimizing the support structure to improve design and construction schemes.