Tunnel boring is a common construction method for urban subway tunnel excavation. However, the tunnel boring process can easily induce ground and nearby building foundation deformations. Electrical transmission towers are crucial infrastructure for urban power supply, and when tunnel boring approaches these towers, an assessment of induced tower foundation settlement should be carried out. The finite element method is a powerful tool for conducting such assessments. However, due to the complexity of the geology and construction conditions, in regions lacking engineering experience and field data, finite element predictions often exhibit significant errors and are challenging to correct. In light of these challenges, this paper proposes a correction method for the finite element prediction model of tower foundation settlement near shield tunnel excavation based on the principle of similarity between ground and foundation settlement in the same finite element model. This method first utilizes pre-measured ground settlement data to determine the expression for its finite element model factor and then introduces it as a correction term into the tower foundation settlement prediction model. The empirical constants in the correction term are updated in real-time using actual tower foundation settlement data obtained during the tunnel boring process, leading to dynamic adjustments and improved accuracy in the prediction model. Using a shield tunnel project in Foshan, Guangdong as an example, this paper validates the superiority and practical value of the proposed method. The research indicates that: Without correction, the finite element model of the tower foundation overestimated its settlement by approximately 30%, with moderate accuracy and variability in predictions. Additionally, the predicted tower inclination based on settlement data did not align with the measured values. After applying the calibration method proposed in this paper, the overall accuracy of the tower foundation finite element model became unbiased, with reduced variability. The inclination of the tower, calculated based on the calibrated settlement data, exhibited high congruence with the measured values and consistent trends.