Multi-objective Optimization Design In Urban Stormwater Drainage System

The process of urbanization leads to a great change to urban land surface and runoff generation and confluence.It also affected the nature hydrologic circulation and resulted in a series of stormwater problems,such as frequent rainstorm,waterlogging,slope instability,etc.The unreasonable design of urban rainwater drainage network and the inadequate stormwater management policies are two important reasons for these problems.Therefore,the design of economical,reasonable and efficient rainwater drainage network and stormwater management plan is of great significance to urban safety and sustainable development.Firstly,the actual condition of hydrometeorology,topographical terrain,planned land use and drainage system in the study area is analyzed,and the SWMM model is established,and the rationality analysis of the model parameters is carried out.For the rainwater drainage network,this paper aimes to reduce the cost and improve the hydraulic performance,the multiobjective optimization model was established with the pipe diameter and slope as optimization variables,and the hydraulic conditions required by the specification as constraints.NSGA-II was chosen to solve the problem,Python was used to embed the SWMM source code into the optimization process,and a heuristic method was developed to obtain a cost-benefit relationship.By comparing the six typical plans with the original one,the cost of the optimized plan decreased by 13.9%,13.2%,12.5%,11.1%,7.4% and 3.8% respectively while meeting the drainage capacity.The optimization technology is proved to be to feasible to be utilized in the design of large-scale rainwater drainage network,and the feasibility and applicability of drainage by pressure flow are explained.For urban stormwater management,this paper establishes a primary selection system for LID facilities based on the low-impact development concept,and finally selects three measures of sunken green area,permeable pavement and green roof to design the study area.Based on the analysis of the suitable area of each measure in the study area,a multi-objective optimization model was established with the objective of reducing the cost of the plan and the total runoff,and the area of three LID measures in each sub-catchment area as variables.The NSGA-II is also used to solve the model,and the cost-benefit balance relationship is obtained.The results show that there is a “cost-effective” high efficiency zone for controlling rainwater runoff by using LID measures.Four typical plans in the “cost-effective” high efficiency zone are evaluated by fuzzy optimization with five different weight combinations,the annual runoff total control rates were 75%,75%,80% and 82%,respectively.It is found that the first plan(LID facilities construction area accounts for 27.3% of the total area)has obvious advantages in cost and the fourth plan(LID facilities construction area accounts for 44.8% of the total area)has obvious advantages in rainwater runoff control.