| In recent years,with the rapid urbanization in China,the construction of the urban flood control project such as grey facilities has caused serious damage to the water ecosystem of the basin and city.In addition,there are some deficiencies in the management system of urban rainwater and flood in China,which aggravate the risk of urban rainwater and flood disaster.Sponge City,which is a new generation of urban planning and construction mode,can effectively improve the management capacity of rainwater and flood,reduce the risk of rainwater and flood disaster in China.Since it was put forward,it has been highly valued by governments at all levels and relevant scholars,and actively carried out practical exploration in various regions.However,there are some problems in the construction of sponge City at present,such as the limitation of sponge design in the city,the inaccuracy of rainwater and flood simulation,and the lack of systematic suitability evaluation in the early stage of planning and construction of sponge city.This has increased a lot of difficulties for the realization of the high-quality sponge city construction objectives and the satisfaction of the security needs for citizens,and has caused certain obstacles.With the development of big data,cloud computing and other related technologies,various kinds of artificial intelligence methods have been applied to urban planning and construction,which provides a new technical means for the intelligent planning and scientific implementation of sponge city.Based on the above-mentioned practical needs and problem cognition,this paper takes Changting County,Fujian Province as an empirical City,selects the rainfall return period of 20-year,and introduces the traditional statistical model and neural network model into the basin-urban rainwater and flood simulation model by combining hydrology and hydrodynamics.From the whole process of rainfall,runoff generation,confluence,suitability evaluation,the spatial adaptability of sponge city system planning was studied in the Changting upstream basin and Changting urban area.Firstly,through the LSTM neural network to train the historical rainfall data of Changting County,we can find the inherent rainfall time series rule,combined with the local rainstorm intensity formula to build the "Changting rain type".Secondly,SCS method and distributed unit hydrograph method are used to simulate the runoff generation process and confluence process in the Changting upstream basin,and the direction of rainwater storage in the Changting upstream basin is proposed from the perspective of "Big Sponge Body" including mountains,forests,fields,lakes and grass,so as to minimize the pressure of rainwater and flood for the Changting urban area.Thirdly,with the help of DHI Mike series software,use Mike Urban to calculate the process of urban pipeline flow movement,use MIKE 11 to calculate the process of river flood movement,and use MIKE 21 to calculate the process of urban surface overland flow.Then,the three models are coupled by various connection methods in Mike Flood to realize the coupling treatment of multiple water movements in the city,and simulate the process of waterlogging in the city.Finally,using python programming to crawl the social and economic data of Changting,combined with remote sensing image data and statistical yearbook data,the paper constructs the construction suitability evaluation system of Changting sponge city from three aspects of rainwater and flood risk,built environment and population density.After calculating each index,the entropy method is used to determine its weight,and the construction suitability of sponge city in Changting city is evaluated based on BP neural network model.It is found that the peak value of 24-hour rainfall pattern of seven rainfall stations related to Changting is mostly between 15-19 hours,and only there are many obvious peaks in Changting Techang station.Under the condition of 20-year rainfall,a large amount of surface flow is produced in the Changting upstream basin,which flows out from Guanyinqiao hydrometric station,causing great pressure of rainwater and flood for Changting urban area.There are many problems in Changting urban area,such as low density of pipe network,small diameter of pipe and silting up of river.The area along the Tingjiang River and in the historical city is the most serious area of water logging.The flood in the Changting upstream basin should be controlled,the construction of sponge body in the Changting upstream basin needs to be promoted simultaneously with the Changting urban area.In the Changting urban area,there ara good sponge city construction suitability along the Tingjiang River,on both sides of Zhaozheng Road,Shuidong street historical and cultural block,some parks and residential land,which are high risk of waterlogging and large population density,so these areas are the key point of sponge city construction.It should be determined as the near-term construction scope of Changting sponge city construction.The research of this paper shows that by using the method of artificial intelligence,the basin and city are combined to consider flood control as a whole,the surface overland flow,pipeline flow and river flow inside the urban area are combined to simulate the process of rainwater and flood,and the construction suitability of sponge city is evaluated comprehensively by combining the economic,social and ecological environment.It has important guiding significance for the determination of construction time sequence,the division of spatial pattern and the selection of sponge body in the sponge city construction of Changting.At the same time,a series of simulation analysis methods constructed in this paper have relatively low requirements for data,and have strong applicability for sponge city planning of small and medium-sized mountain cities.The whole thesis contains about 45 000 words,45 pictures and charts. |
PDF Full Text Request