In spite of the developments in the application of evolutionary algorithms in the urban water sector, it is not widely used by practitioners in their day to day work. A reason for this underutilization is the lack of accessible optimization tools. The available limited options, on the one extreme, are characterized by tools for very specific applications (e.g. pipe sizing): While these provide useful optimization applications in specific domains, they are too restrictive in the range of application. The other extreme is to `glue-together’ a general purpose EA library and other required algorithms (e.g. an urban drainage simulation model) using computer programming languages. While this approach provides flexibility to potentially implement any optimization scheme, the computer programming skills demanded from the user make it inaccessible for many. A software tool was developed to help urban water engineers to lean the application of evolutionary computing techniques is presented. A popular urban drainage network modeller is coupled with an evolutionary computing library to create an optimization system driven by an accessible graphical user interface. The system is implemented in Python language using free and open-source tools and is released under a permissive licence. The design approach results in a software-tool that does not sacrifice range of applicability while being user-friendly. The tool was tested out in a number of graduate school classes and found to be effective in helping students internalize the principals of EA and its application in urban drainage/sewerage sector.
Pathirana, Assela, "SWMM5-EA – A Tool For Learning Optimization Of Urban Drainage And Sewerage Systems With Genetic Algorithms" (2014). CUNY Academic Works.