Skip to content
@SWMMBobSWMM6

SWMM6 Enablement

Where EPA SWMM goes next: transient flow, reinforcement learning, and model coupling

SWMM6 Enablement

Where EPA SWMM goes next. This organization is the research frontier: transient flow, reinforcement learning, model coupling, and the large model corpora needed to test all three.

The model corpus

1729-SWMM5-Models-2030 is a library of 3,800+ EPA SWMM5 .inp models, each with an auto-generated Markdown summary. It spans hydrology, hydraulics, LID, force mains, pumps, weirs, orifices, and dynamic-wave routing, plus XPSWMM conversions and training examples. When you need to prove an engine change holds across thousands of real networks, this is the test set.

Next-generation engines

ITM_SWMM couples Arturo Leon's Illinois Transient Model with SWMM, for the mixed free-surface and pressurized flow that the standard Preissmann slot only approximates.

openswmm.engine and HydroCouple carry the C engine and the interface specifications for coupling SWMM to other models.

Machine learning on the engine

openswmm.gymnasium wraps the SWMM engine as a reinforcement-learning environment. Train a control policy for real-time gate and pump operation against the model instead of hand-tuning rules.

pybme brings Bayesian Maximum Entropy (Serre and Christakos) to Python, for spatial estimation of rainfall and other fields that feed the model.

Browser tools

swmm5-runoff-explorer rebuilds SWMM5 subcatchment runoff in the browser so you can watch the nonlinear reservoir respond to each parameter.

The other accounts

About

Maintained by Robert (Bob) Dickinson, Autodesk Water Technologist and co-developer of EPA SWMM versions 3, 4, 5, and 5+. Fifty-plus years of continuous SWMM work, from punch cards at the University of Florida in 1974 to reinforcement learning on the engine today. Chair of the SWMM5+ Technical Advisory Committee at CIMM.org and member of the EWRI Stormwater Modeling Committee.

Contact: robert.dickinson@gmail.com · X @RDickinson


Experimental work. Verify all results against the production engine before using them for design.

Pinned Loading

  1. 1729-SWMM5-Models-2030 1729-SWMM5-Models-2030 Public

    A library of 3,800+ EPA SWMM5 (.inp) models — hydrology, hydraulics, LID, force mains, pumps, weirs, orifices, and dynamic-wave routing, plus XPSWMM conversions and training examples — each with an…

    Batchfile

Repositories

Showing 10 of 11 repositories

Top languages

Loading…

Most used topics

Loading…