Search

Urban flooding and waterlogging are becoming increasingly serious due to rapid urbanization and climate change. The stormwater management philosophy of low-impact development (LID) has been applied in urban construction to alleviate these problems. The selection and placement of LID designs are the most important tasks. In this study, LID experiments were performed to calibrate the Storm Water Management Model (SWMM). Then, a multi-objective optimization model, which adopted the minimum surface runoff coefficient, surcharge time, and investment cost as objectives, was established by coupling the SWMM and non-dominated sorting genetic algorithm-II (NSGA-II). Hydrological simulations were performed with the SWMM, and optimal calculations were conducted with NSGA-II. Real-coded optimal variables containing detailed size and location information of multiple LID measures were generated, and a decision space for LID design selection was obtained. The optimization designs reduced the surface runoff coefficient from 0.7 to approximately 0.5, the conduit surcharge duration was reduced from 1.62 h to 0.04–0.47 h, and the total investment cost only ranged from 395,000–872,000 ¥. Thus, the optimization model could achieve synchronous optimization of all objectives. This study could provide valuable information for LID design with the aim of urban flooding and waterlogging control. [Display omitted] • Calibrating SWMM parameters based on LID experiments in the area without measured data. • The surcharge status of drainage systems is taken into account in LID optimization. • The structure of optimal variables is suitable for optimization in small-scale communities. [ABSTRACT FROM AUTHOR]