Multiobjective Design and Innovization of Robust Stormwater Management Plans

摘要: In the United States, states are federally mandated to develop watershed management plans mitigate pollution from increased impervious surfaces due land development such as buildings, roadways, and parking lots. These require a major investment in water retention infrastructure, known structural Best Management Practices (BMPs). However, discovery of BMP configurations that simultaneously minimize implementation cost pollutant load is complex problem. While not required by law, an additional challenge find only meet current targets, but also take into consideration anticipated changes future precipitation patterns climate change. this dissertation, multi-scale, multiobjective optimization method presented tackle these three objectives. The demonstrated on Bartlett Brook mixed-used impaired South Burlington, VT. New contributions work include: (A) A for encouraging uniformity spacing along non-dominated front evolutionary optimization. This implemented differential evolution, validated standard benchmark biobjective problems, shown outperform existing methods. (B) procedure use GIS data estimate maximum feasible locations sizes subwatersheds. (C) multi-scale decomposition problem precalculates optimal configuration across entire range possible treatment levels within each subwatershed. one-time pre-computation greatly reduces computation during enables formulation real-valued global optimization, thus permitting evolution. (D) Discovery computationally efficient surrogate sediment load. nine real watersheds with different characteristics used initial stages further reduce computational burden. (E) lexicographic approach incorporating third objective finding solutions robust (F) visualization methods discovering design principles dominated solutions. first simple truss beam problems then provide insights plans. It how applying sensitivity can help one discover uncertain forcing conditions. particular, applied new may make more