A stochastic approach for automatic generation of urban drainage systems

摘要

Typically, performance evaluation of new developed methodologies is based on one or more casenstudies. The investigation of multiple real world case studies is tedious and time consuming.nMoreover extrapolating conclusions from individual investigations to a general basis is arguablenand sometimes even wrong. In this article a stochastic approach is presented to evaluate newndeveloped methodologies on a broader basis. For the approach the Matlab-tool “Case StudynGenerator” is developed which generates a variety of different virtual urban drainage systemsnautomatically using boundary conditions e.g. length of urban drainage system, slope ofncatchment surface, etc. as input. The layout of the sewer system is based on an adaptednGalton-Watson branching process. The sub catchments are allocated considering a digital terrainnmodel. Sewer system components are designed according to standard values. In total, 10,000ndifferent virtual case studies of urban drainage system are generated and simulated.nConsequently, simulation results are evaluated using a performance indicator for surface flooding.nComparison between results of the virtual and two real world case studies indicates the promisenof the method. The novelty of the approach is that it is possible to get more general conclusionsnin contrast to traditional evaluations with few case studies.