Application of Machine Learning in the optimization of the Jet Grouting technique to reduce the impact of deep excavation on the adjacent metro line

Abstract

This paper proposes a new approach for optimizing the structure. The case study implemented an optimization of the Jet Grouting block that is used to reduce the influence of deep excavation on an adjacent Metro. The proposed approach is based on the Multivariate Adaptive Regression Splines (MARS) technique and Multi-Objective Genetic Algorithm (MOGA). The investigated data was adopted from previous research that analyzed the effectiveness of the Jet Grouting technique in reducing the influence of deep excavation on the adjacent Metro. The analysis results show that the suggested size of Jet Grouting from the proposed approach can effectively optimize both lateral displacement and construction cost. The proposed approach can be applied to the optimization of other civil engineering problems.