The effective and efficient optimization of Storm Water Management Model (SWMM) parameters is critical to improving the accuracy of the urban rainfall-runoff simulation. Therefore, it is necessary to investigate the a...
详细信息
The effective and efficient optimization of Storm Water Management Model (SWMM) parameters is critical to improving the accuracy of the urban rainfall-runoff simulation. Therefore, it is necessary to investigate the applicability of the dynamic system response curve (DSRC) method in optimizing SWMM model parameters, which is newly proposed to solve the nonlinear problems encountered by current widely used optimization methods. A synthetic case, free of data and model errors, was used to examine the applicability of the DSRC with single-objective or multi-objective functions in finding the optimum parameter values known by assumption. A real watershed case was selected for the optimization of SWMM parameters by use of DSRC with the most suitable objective function, which was determined by a synthetic case. In addition, the advantages of the DSRC in SWMM parameter optimization over the Particle Swarm Optimization(pso) and Multiple Objective Particle Swarm Optimization(mopso) algorithms were analyzed in terms of NSE, REv, REp, and EPt. The results revealed that the DSRC with multi-objective function could find the global optima of all SWMM model parameters in the synthetic case, but it could only attain part of them with a single-objective function. In the real watershed case, the DSRCS-optimized SWMM performed better than mopso-optimized one with an increase of average NSE by 5.8% and a reduction of average vertical bar REv vertical bar, vertical bar REp vertical bar and vertical bar EPt vertical bar by -53.7%, -67.9%, and -34.6% respectively during the study period. The outputs of this paper could provide a promising approach for the optimization of SWMM parameters and the improvement of urban flooding simulation accuracy, and a scientific support for urban flood risk control and mitigation.
暂无评论