Study On Real-time Control Model Of Urban Drainage System In One River Catchment Of Guangzhou City

Urban water-logging occurs frequently in China,which causes great social and economic harm.Urban water-logging remote monitoring and control system is one of main measures to reduce urban water-logging,but it is difficult to be popularized and applied due to its single function,difficulty in use and high cost.This study attempts to mitigate urban water-logging risk through establishing real-time monitoring cloud service system and urban drainage system real-time control model,which has the characteristics of efficient,simple and low-cost management of urban water-logging risk.One river watershed of Guangzhou city is taken as the studied area,urban water-logging real-time monitor system was established,and on-site installation of system terminal equipment and the test of system-running were completed.The results show that the system runs stably,and real-time data can be collected and uploaded timely and accurately.Then,based on Environmental Protection Agency Storm Water Management Model(EPA-SWMM)software,we have finished the generalization of river channel and drainage pipelines,the addition of auxiliary facilities of urban drainage system and the division of sub-catchment areas.The model parameters calibration of the model and the validation of the model are tested through six rainfall events with senors-measured nodes data respectively.The results show that the selection of the model parameters is reliable and the UDS model is effective.Finally,particle swarm optimization(PSO)algorithm is used to optimize the real-time control operation of UDS.We finished the selection of PSO algorithm parameters.And then,a new running-mode of real-time control model of UDS based on real-time sensor data is proposed.The results of real-time control of three rainfall events in low tide,high tide and storm surge respectively show that the total overflow decreased 37.3%,10.8% and 5.4%respectively,the total overflow time of nodes decreased 43.6%,35.4% and 11.4% respectively,and the average lag start time of node overflow lags about 18.1 minute,9.67 minute and 18.78 minute respectively,compared to original control schemes.Therefore,using PSO algorithm to optimize the real-time control operations of urban drainage system and running the real-time operation model in the way mentioned above can effectively mitigate urban flood risk in the watershed;The real-time control results of the model under different rainfall return periodsindicate that when the watershed encounters a rainfall event more than 3-year rainfall return periods,the real-time control effect is limited in the watershed.