This study develops an improved PSO-BP neural network model to predict surface settlement during rectangular pipe jacking tunnel construction. The model, tested on a real-world case study, outperforms traditional methods by accounting for adaptive mutation and incorporating inertia weight and mutation particles for enhanced accuracy and stability. The findings suggest this model can be valuable for similar projects.

The authors acknowledge funding from several sources, including the National Natural Science Foundation of China, the Natural Science Foundation of Hunan Province, and the Science and Technology Innovation Project of Yiyang City. No direct conflicts related to specific companies or products were identified, but the funding sources could potentially influence research directions or interpretations.

The study presents a novel approach to predicting surface settlement during rectangular pipe jacking tunnel construction, using an improved PSO-BP neural network model. The methodology is sound, and the results demonstrate improved accuracy compared to traditional methods. However, the limitations related to generalizability, oversimplification of factors, and limited comparison with other models prevent a higher rating. The funding sources raise a potential for indirect conflict of interest.