Abstract: To achieve precise and effective prediction of shield metro tunnel settlement during construction,ensuring construction and operational safety,this study utilized monitoring data from the Shenyang Metro Line 1 shield tunneling construction site.A Whale Optimization Algorithm (WOA) was introduced to optimize the parameters of a Back Propagation (BP) neural network,establishing a WOA-BP predictive model for subway shield tunnel settlement.A comparative analysis was conducted against three predictive models optimized by the Cuckoo Search (CS) algorithm,Particle Swarm Optimization (PSO),and BP neural network.The results demonstrate that the WOA-BP model for predicting tunnel crown settlement overcomes drawbacks common in traditional settlement prediction models,such as slow convergence speed and a tendency to fall into local minima.This model exhibits superior nonlinear mapping capabilities,enabling efficient and accurate prediction of surrounding rock deformation.The findings provide technical support for intelligent prediction of tunnel settlement and offer valuable guidance for similar engineering projects.