Abstract: To address the issue that current urban flood simulations do not fully consider the impact of pipeline siltation on surface and subsurface runoff generation and confluence, an urban flood simulation model considering the effects of surface and pipeline siltation is developed by integrating the surface-pipeline siltation 1D-2D coupled governing equations, the surface-subsurface retention box model, and the intelligent prediction method for the equivalent siltation degree of pipe networks. Taking Changning District of Shanghai as an example, different measured rainfall scenarios were selected to compare the performance of flood simulation. The results show that: ① Compared with other intelligent prediction algorithms, the intelligent prediction method for the equivalent siltation degree of pipe networks proposed in this study has significant advantages in terms of accuracy and robustness. ② Under measured rainfall conditions, the accuracy of the simulated values of inundation depth and inundation area is significantly better than that of the traditional rainstorm numerical model. The prediction accuracy of inundation area remains above 86.1%, and the mean absolute error of the inundation depth prediction results is reduced by 0.13. ③ When the recurrence period of rainfall is greater than 2 years, the reduction rate of inundation volume under pipeline dredging conditions shows a gradually decreasing trend with the increase of rainfall return period. The reduction rate of inundation area by pipeline dredging presents a trend of first increasing and then decreasing. This study can provide technical support for urban flood simulation and precise disaster prevention and mitigation.