Automatic calibration of SWMM model with adaptive genetic algorithm

Storm Water Management Model (SWMM) is a popular simulation and management tool for sewer system or storm water management. Since it is a physically based model, the calibration process is necessary before a successful implementation. By separating the calibrated parameters into universal and special styles, the shortcoming of ignore differences among subcatchments´ width is conquered, and solution space is also reduced greatly than the way of regarding all calibrated parameters as special parameters. Using flow rate of pipes as objective values, an objective function of difference between simulated results and objective values is build as the objective function of calibration optimal model. A case sewer network is used to evaluate the proposed calibration method, and by comparison with the calibrated results of calibration optimal model using the all universal calibrated parameter selection concept, the advantages of proposed method were summarized.