Research On Parameter Optimization And Application Of Urban Stormwater Pipe Network Model

Under the background of frequent occurrence of urban severe waterlogging disaster and construction of sponge city,urban storm water model has become a kind of effective storm water management means,aiming at contributing advice and suggestions for solving the essence problem of urban storm drainage——the spatial distribution of stormwater.However,when applying,people lack corresponding model calibration,which results in great deviation between simulation result and actual situation.The parameter optimization & calibration is a critical problem of the application accuracy and reliability of urban drainage pipe network simulation model,have been paid increasingly more and more attention in recent years.In the present dissertation,a systematic research was done on automatic optimization of parameters in hydrologic and hydraulic module based on SWMM model which is one-dimensional simulation core of various urban stormwater pipe network model software and optimization theory for parameter identification.The mathematic model of urban stormwater pipeline network and its simulation theory was introduced in details based on SWMM model.Local sensitivity analysis of 14 hydrologic and hydraulic parameters in SWMM was made adopting modified Morris screening method with rainwater pipe network in C city as an example based on studying the hydrological and hydraulic parameter value range of SWMM model systematically.Thus,the sensitivity;distribution of various parameters to different target variables(total runoff volume,peak flow,hydraulic loading at outlet)and the stability analysis result was summarized.Moreover,the set of 10 optimized parameters was finally determined by introducing area and percentage of impervious area correction factor of subcatchment.Adaptive Chaos Particle Swarm Optimization Algorithm(ACPSO)was studied.Standard particle swarm optimization Algorithm(SPSO)was improved by introducing nonlinearly decreasing as per the exponential function adaptive inertia weight and logic self-mapping chaos search technology.Plus,three classical Benchmark problems were adopted for its performance tests,the results show that the algorithm could get rid of premature convergence effectively and strengthen convergence rate and stability.Thus,it can be used for parameter optimization research.Parameter optimization of SWMM model was studied.10 external API functions of SWMM model were developed secondarily adopting C language and used for the research in this paper.Taking water level and flow as the matching object,mathematic model of SWMM parameter optimization was established and it was solved with ACPSO algorithm.Results show that the rainfall used for calibration and verification,Nash–Sutcliffe Efficiency(NSE)Coefficients of two target variables are above 70% and the absolute values of relative percentage error of peak flow are within 10%.At the same time,the verification result of waterlogging overflow is substantially consistent to the actual situation.Thus,the accuracy and reliability of SWMM model after parameter optimization is satisfying.As the application of urban stormwater pipe network simulation model,the operation performance of network in one drainage subarea of C city was assessed based on the above research results.The overall standard-reaching rate of stormwater pipe network was 57.69%(meeting design recurrence interval P=2a and above),but the opposite overall resilience of system reached 98.73%,which indicate that it’s necessary and feasible to develop and construct sponge city in this region.