Developing SWMMCALPY : an automated, genetic approach to calibrating the storm water management model

Parameter calibration is considered a crucial, albeit arduous, step for reliable performance of the Storm Water Management Model (SWMM) that engineers often undertake manually. This research presents an open-source, automated calibration routine that returns a calibrated model input file to the user. The routine first represents the catchment network as a directed graph object using the NetworkX python package for flexibility in handling real-world observed data availability. Once the calibratable subset of the system is identified, a multi-objective, genetic algorithm (modified Non-dominated Sorting Genetic Algorithm II: NSGA-II) estimates the Pareto front for the objective functions within the feasible performance space. The solutions on this Pareto front represent the optimized parameter sets for matching simulated and observed catchment behavior. A specific solution among this Pareto set can be chosen by assigning weights to the objective functions. This solution is then returned to the user, completing a fully automated process that requires minimal user input and does not require intervention or selections during calibration.